Skip to main content

Full text of "Comparative physiology of temperature regulation"

See other formats



: v:~ j: 















Grateful acknowledgement is made to the following for permission to 
publish copyrighted figures: American Physiological Society, George 
Banta Publishing Co., St. Martin's Press, the Missouri Agricultural Ex- 
periment Station, and Reinhold Publisher's. 




Symposium held July 17, 18, 19, 1961 at the 

Arctic Aeromedical Laboratory 

Fort Wainwright, Alaska 

Edited by 
John P. Hannon, Head, Department of Physiology 
Arctic Aeromedical Laboratory 
Eleanor G. Viereck, Research Physiologist 
Arctic Aeromedical Laboratory 

Symposium held under the auspices of 

Geophysical Institute 

University of Alaska 

College, Alaska 


Part I. The Evolution of Homeothermy. 

1. Introductory Remarks, 

John P. Hannon ^^^ 

2. Some Temperature Adaptations of Poikilothermic 

Vertebrates, C. Ladd Prosser 1 

3. The Evolution of Avian Temperature Regulation 

William R. Dawson 45 

4. The Evolution of Mammalian Temperature Regulation, 

Kjell Johansen 73 

Part II. The Maintenance of Homeothermy. 

5. Heterothermy in the Cold Adaptations of Warm 

Blooded Animals, Laurence Irving 133 

6. Maximal Metabolism and Organ Thermogenesis in 

Mammals, Ladislav Jansky 175 

7. The Physiology of Mammalian Cold Acclimation, 

J. Sanford Hart 203 

8. Temperature Responses and Adaptations in Domestic 

Mammals, Max Kleiber 243 

9. Temperature Responses and Adaptations in Birds, 

George C. West 291 

10. Racial Variations in Human Responses to 

Temperature, Frederick A. Milan 325 

11. Temperature Regulation in Animals Native to 

Tropical and High Altitude Environments, Peter R. 
Morrison 381 

12. Temperature Regulation in Animals Native to the 

Desert Environment, Jack W. Hudson 413 



G. Ladd Prosser obtained his B.A. at the University of 
Rochester in 19 29, his Ph.D. at the Johns Hopkins University 
Zoology Department in 1932, and now holds the title of Professor 
of Physiology, University of Illinois. He has performed outstanding 
work in the fieldsofphysiology of nervous systems of invertebrates, 
comparative pharmacology of hearts, conduction in non-striated 
muscles, and physiological adaotation: (1) mechanisms and (2) role 
in evolution. The text. Comparative Animal Physiology, edited 
by Prosser, et al., needs no introduction. 

William R. Dawson received his Ph.D. from Stanford, Depart- 
ment of Zoology, in 1953, and is now assistant professor in the 
Department of Zoology at the University of Michigan. He has worked 
on temperature regulation and water balance of birds and reptiles, 
reptile metabolism, and avian paleontology. 

Kjell Johansen has the degree Gand. Real, from Oslo, Norway. 
He has worked with many kinds of vertebrates and higher inverte- 
brates. He is concerned with physiological phylogeny, and compara- 
tive cardiovascular physiology. He is now atthe University of Oslo, 
Institute for Experimental Medical Research. 

Laurence Irving received the degree of A.B. at Bowdoin in 
1916, the A.M. from Harvard in 1917, and the Ph.D. from Stanford 
in 19 25. He has been awarded honorary degrees also: an honorary 
M.D. from Oslo in 1956 and an honorary D.Sc. from Bowdoin in 
1959. His long and productive list of research accomplishments 
include work in diverse fields such as adaptations to cold in mam- 
mals and birds, migrations of Alaskan birds, and respiratory 
exchange in diving mammals. He has elucidated the principle of 
peripheral heterothermy. 


Ladislav Jansky received the P.H. degree in 1958 from Charles 
University, Prague, £tepartment of Animal Physiology. He now holds 
the title of Assistant Professor, Department of Comparative Phys- 
iology, Charles University. His research interests are: metabolism 
in the cold, cold adaptation, maximal metabolism, organ thermo- 
genesis, and hibernation. 

J. Sanford Hart received the B.A. 1939, the M.A. 1943, the 
Ph.D. 1949 all from Toronto, and was elected a F.R.S.C. in 1956. 
His present title is Head, Environmental Physiology section, 
Division of Applied Biology, NRC, Ottawa, Canada. His general 
interest encompasses theenvironmentalphysiology of homeotherms, 
acclimatization in mammals and birds native to cold climates 
metabolic mechanisms of acclimation to cold in rats, shivering 
mechanisms and organ thermogenesis, and cardiovascular res- 
ponses to exercise and cold exposure. 

Max Kleiber obtained the D.Sci. in 1924 and the Privat Dozent 
in 1928 from the Federal Institute of Technology, Zurich, Switzer- 
land. He is now Professor Emeritus at the University of California 
and has been awarded an honorary LL.D. from California in 1961. 
His numerous contributions to physiology have been in the fields 
of energy metabolism and feed utilization, environmental tempera- 
ture, energy metabolism and food utilization (in chicks and cows), 
measurement of replacement equivalents for food in animal produc- 
tion, energy metabolism and nutritive deficiencies, intermediary 
metabolism especially in milk formation using radiocarbon (C ) 
as a metabolic tracer, and phosphorus and calcium metabolism 
using P and Ca as tracers. His recent book. Fire of Life, is 
destined to become a classic. 

George C. West received his B.A. degree from Middlebury 
College in 1953, and his M.S. and Ph.D. in Zoology under Dr. S. C. 
Kendeigh at the University of Illinois in 1956 and 1958 respectively. 
He spent a year doing post-doctoral research at Illinois and a year 
as a post-doctoral fellow of the National Research Council of 
Canada with Dr. J. S. Hart. He is now Assistant Professor of 


Zoology at the University of Rhode Island and on the summer staff 
of the Canadian National Research Council. His major research 
interests and publications are in the fields of ecology and physio- 
logy of avian energetics and migration. 

Frederick A. Milan obtained his B.A. in 19 52 from the Univer- 
sity of Alaska and his M.A. from the University of Wisconsin, 
1959. He is currently studying for the Ph.D. degree at the Univer- 
sity of Wisconsin, Department of Anthropology. His investigations 
of human response to cold have taken him to the Australian interior, 
the arctic coast of Alaska, the Tierra del Fuego, Lappland, the 
Antarctic, the Canadian Archipelago, and mountain summits in 

Peter Reed Morrison has a B.S. from Swarthmore College in 
1940 and a Ph.D. from Harvard University Biology Department 
in 1947. Presently he is Professor of Physiology and Zoology, 
University of Wisconsin. His work includes the fieMs of tempera- 
ture regulation and energy metabolism in wild mammals, altitudinal 
adaptations in wild mammals, coagulation and the polymerization 
of fibrinogen, hibernation, physiological descriptions of intact 
animals, and individual tissue specializations. 

Jack W. Hudson obtained his M.A, from Occidental College, 
Biology Department, 1950 and his Ph.D. from University of 
California, Los Angeles, Zoology Department, 1960. He is now 
a lecturer in that department. The list of his research interests 
is.: Aestivation and hibernation in mammals, water metabolism and 
kidney function in desert mammals, uric acid excretion in birds, 
relation of heart rate, body size and thermal neutrality in birds. 


Adams, Thomas 
Department of Physiology 
University of Washington 

Hannon, John P. 
Physiology Department 
Arctic Aeromedical Laboratory 
Fort Wainwright, Alaska 

Durrer, John P. 
Physiology Department 
Arctic Aeromedical Laboratory 
Fort Wainwright, Alaska 

Kessel, Brina 
Department of Biology 
University of Alaska 
College, Alaska 

Eagan, Charles J. 
Physiology Department 
Arctic Aeromedical Laboratory 
Fort Wainwright, Alaska 

Evonuk, Eugene 
Physiology Department 
Arctic Aeromedical Laboratory 
Fort Wainwright, Alaska 

Folk, G. Edgar, Jr. 
Department of Physiology 
University of Iowa 
Iowa City, Iowa 

Miller, Keith 

Arctic Public Health Research 

P. O. Box 960 
Anchorage, Alaska 

Opsahl, James F. 
Department of Zoology 
Winona State College 
Winona, Minnesota 

Pitelka, Frank 

Museum of Vertebrate Zoology 

Berkeley 4, California 


John P. Harmon 

On behalf of the Commander and Staff of the Arctic Aeromedical 
Laboratory, it is indeed a pleasure to welcome such distinguished 
guests to this second Symposium on Arctic Biology and Medicine. We 
hope that your trip was an enjoyable one, despite the long distances 
many of you were forced to travel. It is our desire that you remem- 
ber this visit to our Laboratory and to the State of Alaska as a 
pleasant one as well as a scientifically profitable experience, and 
if there is anything that I or the other Staff members might do to 
assure this end, please do not hesitate to call on us. Before we 
begin the formal portions of our program, I would like to say a few 
words regarding our reasons for selecting the Comparative Phys- 
iology of Vertebrate Temperature Regulation as a symposium topic. 

As you are all well aware, the ability to adapt to an adverse or 
unusual environment is one of the more fundamental characteristics 
of all living things. In fact, we might go so far as to say that the 
ability to adapt to such environments is an essential prerequisite 
to the successful perpetuationof any population of plants or animals. 
Thus, when a species is unable to adapt to an adverse environment 
it becomes extinct. 

To those ofus wholivein arctic or subarctic areas, the adapta- 
tions of plants and animals to adverse environmental temperatures 
are of singular importance. The accumulation of knowledge about 
such adaptations therefore, is one of the primary reasons for invit- 
ing you to participate in this Symposium. 

Since temperature adaptation is a very broad subject, it was 
obviously impractical to attempt to organize a symposium that would 
adequately cover the whole field. Consequently, we decided to con- 
tinue the pattern that was followed in our first Symposium on Arctic 
Biology and Medicine; namely, to give intensive consideration to 
one rather narrowaspectof this field. Furthermore, we also decided 


that we should give primary emphasis to subject material that had 
not been discussed in detail at previous symposia. 

The Comparative Physiology of Vertebrate Temperature Regu- 
lation seemed to admirably meet these criteria. Here was a sub- 
ject where the work covered practically the whole range of zoologi- 
cal sciences. Here, also, was a subject where the quite similar 
work by investigators in one scientific discipline often went unrecog- 
nized by investigators in another closely related scientific discipline. 
For example, those of us who study the biochemistry of cold accli- 
matization in small laboratory mammals may be unfamiliar with 
biochemical studies on fish or other vertebrate heterotherms; those 
of us who study the temperature regulation of cats and dogs may 
not be cognizant of temperature regulation studies on domestic ani- 
mals such as the cow; and those of us who are concerned with the 
natural temperature adaptations of arctic animals may not relate 
our information to studies that have been conducted on desert 

Our first concern, therefore, in organizing this symposium was 
to bring togetherrepresentativesof the various scientific disciplines 
who are interested in vertebrate temperature regulation. Beyond 
this, we had two other desires in organizing this symposium. One 
of these was to obtain participants who could discuss vertebrate 
temperature regulation from the evolutionary standpoint. Our other 
desire was to obtain participants who could interrelate temperature 
adaptations to other forms of environmental adaptation. I feel that 
we have been at least partially successful in achieving both of these 
desires. And in this regard I would like to express my gratitude to 
Dr. Laurence Irving, Dr. J. Sanford Hart and Dr. C. Ladd Prosser 
for their valuable suggestions regarding possible participants and 
subject material. 



C. L. Prosser 

Temperature limits thedistributionof many poikilothermic ani- 
mals, and knowledge of responses to temperature is important for 
physiological ecology. Natural selection acts on the capacity for 
change within a given genotyTpe; hence it is important to learn how 
such an environmental variable as temperature brings about bio- 
chemical changes in individual animals. Natural variation in respect 
to. temperature relations can best be described in terms of the re- 
sponses to the stresses of cold and heat: survival, reproduction, 
various rate functions, behavior. Once this variation is described 
for natural populations of animals, it is necessary to analyze that 
component which is genetic and that which is environmentally in- 
duced; this analysis is permitted by acclimation of similar animals 
to a range of temperatures. Finally, the physiological mechanisms 
of the variation with respect to temperature can be pursued down to 
the molecular changes, and the sequence of events by which tem- 
perature brings about change in genetically similar individuals can 
be elucidated. 


The differences between a poikilotherm (temperature conform- 
er) and a homeotherm (temperature regulator) are multiple and 
fundamental. Birds and mammals evolved from reptiles and differ 
from present-day reptiles in possession of a thermoregulating center 
in the brain, in insulation, in peripheral vascular responses to ambi- 
ent temperature (which are opposite to those of reptiles), and in type 
of metabolic compensation. Varying degrees of homeothermy ex- 
pressed in hibernation, estivation, nocturnal temperature drop, and 

heterothermy of tissues indicate that some animals can shift from 

homeothermy to poikilothermy and that certain peripheral tissues of 


some birds and mammals can function over a much wider range of 
temperature than can the core tissues. The corresponding enzymes 
must differ in cold functional skin and in constantly warm liver. 

When the ambient temperature (air or water) falls, a homeo- 
therm shows a typical sequence of protective responses. The meta- 
bolic response in relation to body temperature is diagrammed in 
Figure 1. Peripheral cold receptors signal the drop in skin tempera- 
ture and initiate reflexes such as hair or feather erection, peripheral 
vasoconstriction, and behavior such as huddling. These initial re- 
sponses result inheatretentionby increased insulation. With further 
chilling, the temperature- sensitive center in the hypothalamus is 
stimulated and further defenses maybe mobilized. Cooling below the 
critical ambient temperature or to that temperature below which in- 
sulative changes are inadequate so that a transient drop in body tem- 
perature occurs, results in increased metabolism which serves to 
maintain body temperature. Nor- adrenaline secretion is enhanced, 
and shivering maybe initiated and heat production increased. If cold 
stress continues, the hypothalamus activates the anterior pituitary 
to liberate adrenocorticotropic and thyrotropic hormones. The ad- 
renal cortex and thyroid initiate a metabolic increase and extensive 
biochemical responses of various organs, particularly liver to vary- 
ing extents in different species. In laboratory acclimation to cold 
some animals show an increased standard metabolism; in field ac- 
climatization many animals show a reduction in critical temperature. 
Some metabolic enzymes become more active than others and the 
sensitivity to stimulating hormones is altered (Hannon, 1960; Hart, 
1957; Heroux, 1960). The increased metabolism of cold acclimation 
may persist after withdrawal of hormonal stimulation, and in the 
annual cycle of winter, insulative adaptations make metabolic ones 
less necessary. In some species the adrenal cortex is active and 
the thyroid less active under natural winter conditions. 

Over a thermoneutral zone the insulative changes are sufficient 
to maintain relativeconstancyofbody temperature in a homeotherm. 
At elevated ambient temperatures, reflexes provide increased peri- 
pheral bloodfloWjSurfacecoolingbysweat, panting, and other means 
of controlling body temperature. However, there is no reduction in 
metabolism and in conditions of fever the oxygen consumption may 






Body temp. 

c O2 Consumption 


Figure 1. Schematic representation of temperature regulation in a homeotherm. 
At a critical ambient temperature the body temperature is temporarily reduced, but 
metabolism increases, thus maintaining body temperature. C, cold acclimated; W, 
warm acclimated. 


When environmental temperature drops, the temperature of a 
poikilotherm (e. g., fish) drops with it. Any metabolizing organism 
produces some heat, and the liver of a large fish may be significantly 
warmer than its environment. But poikilotherms lack insulation, and 
their body temperatures are virtually the same as that of their en- 
vironments. If the drop in temperature is rapid and considerable, the 
poikilotherm may enter a chill coma and even die from respiratory 
failure. If the cold stress is less, there may be initial stimulation, 
increased nervous activity (well shown in crustaceans and insects) 
and an initial transient increase in oxygen consumption, the so-called 
initial shock reaction. This is followed by a decline of metabolism to 
a stabilized state which corresponds to the reduced temperature. The 
Q for metabolism is usually between 2.0 and 2.5; hence the meta- 
bolic response to temperature is steeper than the change in body 
temperature. With time (days or weeks) some metabolic compensa- 
tion may occur. The compensatory changes for either a fall or a rise 
in temperature in a poikilotherm are diagrammed in Figure 2. The 
time course of acclimation differs according to the function mea- 
sured and the kind of animal. 

Precht (Precht, 1958;Prechtetal., 19 55) has classified the pat- 
terns of acclimation as indicated in Figure 3 and has termed them 
capacity adaptations. The five possible patterns are: (1) overcom- 
pensation so that metabolism is higher in the cold than at the initial 
temperature, (2) perfect compensation with the same metabolism at 
each temperature, (3) partial compensation, (4) no compensation, the 
metabolism continuing to follow the van't Hoff relation, and (5) in- 
verse compensation or further reduction in metabolism. The com- 
monest pattern of acclimation is the third, partial compensation, so 
that if the metabolism of animals from temperatures t and t is 
measured at the same intermediate temperature, the one acclimated 
to the cold has a higher metabolism. This acclimation pattern can 
apply to other rate functions besides metabolism- heart rate, breath- 
ing rate etc. A comparable sequence is described for moderate in- 
creases in temperature (Precht, 1958). Acclimation to heat is a re- 
duction in metabolism below the initial level determined by the Q 

relation (Gelineo, 1959). The net effect of long-term acclimation is 

to tend toward relative constancy of energy liberation despite 
changes in body temperature. 







Figure 2. Schematic representation of temperature relations In a poikilotherm. 
Metabolism decreases more steeply than body temperature. Acclimation results in 
a rise in metabolism at lowtemperatureanda fall at high temperature, thus tending 
toward relative constancy of metabolism as the environmental temperature changes. 



Figure 3.Precht'spattemsof metabolic acclimation in cold. Animal moved from 
a high temperature (tj to a low one (t) and metabolic rate falls directly along solid 
line. Rate remains at 4 ifno acclimation occurs with time (van't Hcff approximation); 
rate rises to 2 if acclimation is complete. Pattern 1 represents over-compensation, 
3, partial compensation, and 5, under-compensation or reverse acclimation. Modi- 
fied from Precht, 1958. 


Tolerance of sudden temperature stress, called "resistance 
acclimation" by Precht, is also modified. This is shown by shifts of 
both high lethal and low lethal temperatures according to acclimation 
(Fry et al., 1946; Fry, 1947). The curves describing rise or fall of 
the two lethal temperatures as a function of acclimation need not be 
parallel, and the area enclosed by both curves, the tolerance zone, 
is species specific. The relation of temperature tolerance or resist- 
ance acclimation to capacity acclimation is not known, and further 
knowledge of heat and cold death might indicate which processes are 
altered. Stress tests provide a useful tool for analysis of acclima- 

Poikilothermic vertebrates differ from homeotherms in that 
they tend by compensation to maintain relatively similar activity 
when body temperature changes, whereas the homeotherm maintains 
constant temperature. There is no "comfort?' or thermoneutral zone 
for the poikilotherm as long as chill or heat coma are avoided. Effi- 
ciency of feeding and general body activity increase in a non- linear 
fashion up to some "optimal" temperature which may be only a few 
degrees below the lethal point. Furthermore, there is no evidence in 
poikilotherms for a sequence comparable to Selye's stress syndrome 
of mammals. 

Whether or not hormones are involved in the enzymatic changes 
of metabolic acclimation in poikilotherms is not known. Evidence 
concerning thyroid participation in adaptation of fish is conflicting 
(Hoar, 1959). A slight increase in height of thyroid epithelium at ele- 
vated but not at reduced temperatures was reported for the minnow 
(Phoxinus) (Harrington and Matty, 1954), and in trout the thyroid 
shows signs of increased activity in the cold (Olivereau, 1955b). 
However, no histological change was found in thyroids of catfish, 
carp, tench, eel, Mugil, or Scyllium after acclimation in cold (7 C- 
14° C) or warm (20 C-23.5 C) (Olivereau, 19 55a,b,c). Thiourea 
treatment is said to eliminate metabolic differences between cold- 
and warm- acclimated crucian carp (Garassiug (Suhrman, 19 55), but 
thiourea increases the differences in Leuciscus (Auerbach, 1957), 
Resistance to cold in long-day goldfish increases when thyroid hor- 
mone is injected (Hoar, 1959). Thiourea increases cold resistance 
of goldfish and decreases that of the crucian carp ( Carassius ) 
(Precht, 1958). Iodine uptake by the thyroid is slightly increased by 


cold in a minnow ( Umbra ) but not in Fundulus (Berg et al., 1959). 
Temperature effect on the thyroid of amphibians is negligible. How- 
ever, seasonal variations in thyroid activity of both fish and amphi- 
bians, probably associated with photoperiod, are considerable. The 
adrenals of poikilothermic vertebrates produce corticosteroids 
which seem to function primarily in potassium and sodium balance; 
no role in carbohydrate metabolism has been found in poikilotherms 
(Jones etal., 1959). The amount of hydroxycorticosteroid in the blood 
of a fish may be increased after swimming but no response to tem- 
perature stress has been reported (Jones et al., 1959). Changes in 
metabolic enzymes in compensation for temperature occur in yeast 
(Precht, 1956) and in invertebrates where thyroxin and iodinated 
tyrosines and corticosterone do not function as they do in homeo- 
therms. Also, the biochemical changes of poikilotherms in tem- 
perature adaptation can be either an increase or a decrease in 
specific enzymes. It seems likely that the acclimation of poikilo- 
therms is either a direct effect of temperature on enzyme forming 
systems or an indirect enzyme induction due to differential utiliza- 
tion of substrates at different temperatures. Thus there is little 
similarity in the metabolic acclimationof poikilotherms and homeo- 


The analysis of biochemical mechanisms of temperature accli- 
mation is beset with many difficulties. The identification of limiting 
steps involves extrapolation to the intact animal from measurements 
on tissue slices, homogenates, isolated mitochondria, and purified 
enzymes. Such extrapolation is difficult and based on several as- 
sumptions. It is not possible to provide in vitro conditions which dup- 
licate in all respects those under which an enzyme functions in vivo . 
Balance of organic as well as inorganic ions, concentrations of co- 
factors and hormones cannotbeduplicated, nor can spatial organiza- 
tion, as of particulates in a cell. An important part of acclimation 
involves regulation by the neuroendocrine system. Yet the integrated 
system can be analyzed only by taking it apart. One method of 



identifying changes in enzyme activity is to observe effects of in- 
hibtors; yet these are not nearly so specific as desired. Another 
method is to purify enzymes, but extraction precedures are often un- 
certain as to recovery or loss of activity. A common method is to 
supply an excess of a specific substrate so that the enzyme acting on 
it is made limiting; this provides a useful comparison between sys- 
tems treated differently (as by temperature) , but it does not tell 
much about limiting steps in vivo . Tracing labelled substrates is in- 
formative and has not often been used in acclimation biochemistry, 
although it has indicated a general similarity of metabolic paths in 
fish and mammals (Brown, 1960; Brown and Tappel, 1959; Martin 
and Tarr, 1961). Useful information can be obtained from kinetic 
studies of both intact and dissected systems. 

A serious problem, especially with poikilotherms, is the identi- 
fication of appropriate environmental variables. Three factors, tem- 
perature, nutrition, and photoperiod, interact in an inextricable way. 
Many fish and amphibia eat little in the cold, and it has been common 
practice to observe acclimation in starved animals. Unfortunately a 
fish starved at 25 C is not comparable in its food reserves to one 
starved at 5 G. Also if each is fed ad libitum , the absorption of food 
may be so slow in the cold that the nutritional state is different from 
that of one fed at 25 C. We have evidence that the metabolic differ- 
ences are greater in starved than in fed goldfish kept at low and high 
temperatures. Various methods, such as feeding followed by cross 
acclimation so that the total time spent at the two temperatures is 
the same for both groups, have been used in an effort to approach 
nutritional equivalence, but no method is fully satisfactory. 

Photoperiod has marked metabolic effect in fish and amphibians. 
Ekberg (1961) found a greater difference between enzymes from cold 
and warm acclimated fish on a 17-hour than on a 7-hour photoperiod; 
he also found a marked seasonal difference in the metabolic response 
of goldfish gills. Roberts (1961) observed a photoperiod effect 
on Carassius carassius at 20 C but not at lower temperatures. 
Hoar (1955; Hoar and Robertson, 1959) observed seasonal dif- 
ferences in temperature tolerance and in oxygen consumption 
by goldfish even when acclimated at the same temperature; 
these seasonal effects reflect photoperiod and may be associated 
with enhanced thyroid activity on short photoperiod. Frogs show 
marked seasonal differences in many of their physiological proper- 



ties independent of temperature. It is important, therefore, that pho- 
toperiod be kept the same for different conditions of temperature 

Another difficulty in metabolic acclimation results from the 
differences between active and rest (standard) metabolism and the 
impossibility of controlling movement in poikilotherms. There is 
evidence that active and rest metabolism follow slightly different 
enzyme pathways. Data from anesthetized fish differ from those 
from quiescent awake ones; hence anesthetics are usually avoided. 

Kinetic analyses have been useful in studies of enzyme induction 
and of the role of amino acid pools in protein synthesis, but such a- 
nalyses have not often been applied to problems of acclimation. 

The time -course of acclimation deserves more attention. One 
related method is to compare the rate- temperature curves of stabil- 
ized rate functions for poikilothermic animals that have been differ- 
ently acclimated (Prosser, 1958). Such curves permit some specula- 
tion concerning themechanismof acclimation (Figure 4). When there 
is no acclimation, the rate-temperature curves coincide for animals 
from either temperature (Figure 4a). This lack of acclimation has 
been described for winter and summer Gunner (Haugaard and Irving, 
1943) and for a variety of insects and shore invertebrates. One t)T)e 
of acclimation to cold is a translation of the rate curve to the left or 
upward (Figure 4b) without change in slope. Such simple translation 
has been observed for O consumption by the scorpene trout (Gelin- 
eo, 1959), cocarboxylaseoftheeel(Carlsen,1953),oxygen consump- 
tion by salamanders, Eurycea (Vernberg, 1952) and Triturus (Riech 
et al., 1960), for metabolism of some northern and southern species 
of frogs (Tashian, 1957), of the lizard Sceloporus at 16 C and 23 C 
(Dawson and Bartholomew, 1956) as well as for numerous inverte- 
brates (Prosser, 1961). 

A third pattern (Figure 4c) is rotation about a midpoint, i. e., 
change in slope or Q only. This occurs for O consumption by the 
European eel with an intersection of curves for 11 C and 26 C 
acclimation at about 21° C (Precht, 1951) and also for metabolism 
by the salamander Plethedon (Vernberg, 19 52). The most common 
pattern is a combination of translation with rotation. When the Q 
of cold- acclimated animals is less than that of warm- acclimated 



0. z 

o g 


UJ _i 
I- o 



\ \ 


\ \ 


\ \ 

^ \ \i 


5 \ \ 

^^ ^^ \ 


^^ X \ 

s \X 


* . ^"^S. 

1 ^ 












\ \ 


<o\ \ 


\ ^ 

s. ~ 


1 N 


3iVa 901 

3iVa 90"l 


§ ^ 

u U 

^ oj 

i 1 



ones, the two curves may intersect by extrapolation above the normal 

temperature range (Figure 4d). Examples for vertebrates are heart 

rate of the newt Triton (Mellanby, 1940), metabolism of cottid fish, 

winter and summer, northern and southern latitudes (Morris, 1961) , 

o o 

and O consumption by frogs acclimated to 5 C and 25 C (Riech et 

al., 1960). If the Q of cold- acclimated animals is higher (Figure 
4e), the two curves may intersect at alow temperature, often by ex- 
trapolation. Above the intersection the rate is greater for cold- ac- 
climated than for warm-acclimated animals. This is reported for 

o o 

O consumption by the crucian carp acclimated to 5 C and 26 C 

(Suhrman, 1955) and for O consumption by brain tissue of goldfish 
(Freeman, 1950). 

Translation of a rate-temperature curve implies a change in ac- 
tivity (in a thermodynamic sense) and may be caused by change in 
enzyme concentration, change in the relative activities of enzymes in 

series or in parallel, or a change in controlling conditions ionic 

strength, pH etc. Rotation of a rate- temperature curve implies a 
change in Q and hence in activation energy and may result from 
alteration of me enzymatic protein, change in some co- factor, or a 
shift in control of a reaction to alternate enzymatic pathways. Differ- 
ent tissues of the same animal may show different patterns of meta- 
bolic acclimation, e. g., the heart of goldfish shows no change, but 
skeletal muscle, and to a lesser degree liver, shows acclimation 
with a reduction of Q in the cold. 

Metabolic and Enzymatic Changes 

A number of selected examples of compensatory acclimation of 
metabolism of intact poikilo thermic vertebrates is given in Table I. 
A greater oxygenconsumptionof cold- than of warm- acclimated ani- 
mals when measured at intermediate temperatures is indicated for 
lampreys (Scherbakov, 1937), eels (Precht, 1951), marine fish Fun- 
dulus and Gellichthys (Wells, 1935a,b), goldfish for both active and 
standard metabolism (Kanungo and Prosser, 1959a) , and frogs (Riech 
1960). The extent to which photoperiod and nutritional state modify 
these differences is not clear, but the principal experimental vari- 
able in each experimentwas temperature. When measured at the ac- 
climation temperatures, the maximum active metabolism is lower 
than the maximum standard metabolism (Fry and Hart, 1948b); if 





Acclimation Temperature 

and Oxygen Consumption 

per Wet Weight 

Per Cent by which 
Cold Exceeds Warm 

(Scherbakov, 1937) 

1.5-3.5° 15-17° 

16° 0.21 mgOg/g^/hr 0.14 



(Preoht, 1951) 

11° 26° 

12° 5 mlOj/lOOgw/hr 2 150% 

22° V mlOj/lOOg^^/hr 7 0% 

Carassius gibello 

crucian carp 10-11 

(Suhrmann, 1955) 20-21^^ 

29 '^ 

4.4 mlOg/lOOgw/br 3-2 

15.5 mlOg/lOOg^^r/hr 6.8 

24.6 mlOg/lOOgw/lir 16.1 




Carassius carassius 

carassius 20' 

(Roberts, 1960) 

70 mlOg/kg/hr 
72 mlOg/kg/hr 

168 (7 hr day) 
238 (17 hr day) 


g°l'*^^^^ Standard f' 

(Kanungo, Prosser, 25 

19 59 a) „„c 

10° 30° 

8.7 mlOg/lOOgw/hr 6.94 

13.7 mlOj/lOOgw/hr 7.86 

11.29 mlO2/l00g^/hr 9.76 

19.33 mlO2/l00gw/hr 12-2 



(WeUs, 1935) 

0.11 gms/hr 




(Riech et al., 1960) 

5° 25° 

10° 8 mlOg/lOOgw/hr 3-0 166% 

20° 11 mlOg/lOOgw/hr 7.5 47% 

(Jankowsky, 1960) 


158 mm3/hr/g2/3 




Table I. Metabolic acclimation of intact poikilothermic vertebrates. 



metabolism of animals acclimated to extreme temperatures is mea- 
sured over the entire curve, the two maxima are similar (Kanungo 
and Prosser, 1959a) (Figure 5). Species differences maybe marked 
as between Carassius carassius and C. gibelio (Suhrman, 1955 and 
Roberts, 1960) ; G. carassius shows an inverse or type 5 acclimation, 
_C. gibelio a positive or type 3 acclimation. 

Data for oxygen consumption by tissue slices, homogenates, and 
whole gills are given in Table II. There is some disagreement for 
the same tissue among investigators, and some tissues show more 
temperature compensation than do others. Ingeneral, skeletal mus- 
cle shows more change than does liver or heart. For brain, compen- 
sation is reported by two authors and lack of compensation by two. 
Gills of fish show marked metabolic compensation. For both gills and 
muscle, Roberts (1960) reports higher daytime metabolic rates and 
slightly greater differences between warm- and cold- acclimated 
tissues when the fish have been on a short day (7-hour) photoperiod 
than on a long day (17-hour) photoperiod. No attempt has been made 
to equate the O consumption by various tissues to the total by the 
intact animal arid to evaluate the relative contributions of each, but 
the percentage of change found for isolated tissues is less than for 
intact animals. 

Evidence for acclimatory effects on some enzymes of poikilo- 
thermic vertebrates and not on other enzymes is summarized in 
Table IE. The reported enzymatic effects are insufficient to account 
for the observed changes in metabolism. Differences in some de- 
hydrogenases and electron transport enzymes have been reported. 
Gills from cold- acclimated goldfish were more sensitive to cyanide 
(Ekberg, 1958) (and liver more sensitive to an timycin (Kanungo and 
Prosser, 1959b)), while liver showed no significant differences with 
respect to inhibition by cyanide, azide, carbon monoxide, or amytol 
(Kanungo and Prosser, 19 59b) . Cocarboxylase showed some compen- 
sation in liver and questionable effect in muscle (Garlsen, 1953). 
Succinic dehydrogenase of eel liver as measured by methylene blue 
reduction showed considerable change (Precht, 1951). In goldfish 
this enzyme was altered in muscle and in liver when measured on a 
protein (but not on a wet weight) basis (Murphy, 1961). Malic dehy- 
drogenase of goldfish liver showed considerable inverse acclima- 
tion (Precht's type 5) (Murphy, 1961). Cytochrome oxidase showed 







.6 - 

.4 - 


/ / / 


Wa Af / 


/ Cs = Standard, 10° 
^s^ Ca = Active, 10° 

Ws= Standard, 30° 

Wa = Active, 30° 

1 1 1 1 1 ._J 

10 20 30 


Figure 5. Oxygen consumption of goldfish measured at different temperatures. 
Both active and standard metabolism offish acclimated at 10 C and at 30 C. From 
Kanungo and Prosser, 1959a (Fig. 1, p. 261). 



Temperature Acclimation Temperature Per Cent by 

Tissue of and O2 Consumption per which cold 

and animal Measurement Unit Weight Wet (w) or Dry (d) Exceeds Warm 

Crucian carp 

(Roberta, 1960) 20"^ 

4-7° (7 hr day) 
290 mnfi/g^Air 
4-7° (17 hr day) 

20" (7 hr) 



(Roberts, 1961) 

0.073 fjl02/mgcj/hr 



(Murphy, 1961) 

20° succinate 

20 glucose 

5"^ (12 hr day) 
652 Ml02/mgw/'"" 

(6-8 days) 
614 (23-35 days) 
538 (6-8 days) 
667 (30-35 days) 




frog (R. pipiens) 
(Riech et al., 1960) 

frog (R. temporaria) 


180 mm^/g^Ar 

440 mm /g/hr 




(Ekberg, 1958) 

Crucian carp 
(Roberts, 1960) 

10° (Feb.) 
18° " 
26° " 





0.257 Ml02/nig<j/hr 



0.557 Ml02/nig(j/hr 



1.074 Ml02/nigd/hr 



4-7° (7 hr day) 

20° (7 hr day) 


701 mm3/g^^,/hr 


4-7° (17 hr day) 

20° (17 hr) 




Table II. Oxygen consumption by tissues (usually with glucose) from poikilo- 
thermic vertebrates acclimated at different temperatures, (n.s., not significant) 



Per Cent by 



and O2 Consumption per 


and Animal 


Unit Weight Wet (w) 

or Dry (d) 






(EWaerg, 1958) 


0.621 Ml02/nigd/hr 














(Kanungo, Prosser 


1.2 >il02/mg(j/hr 








(Murphy, 1961) 


363 ^jlOg/g^/hr 




697 ^102/g^/hr 






(Riechet al., 1960) 









(Ploberts, 1961) 


0.519 (i 102/mgd/l"' 




.(Freeman, 1950) 


9 Mg02/g^/min 






(Ekberg, 1958) 


1.53 lil/mg^Aa 







(Murphy, 1961) 


530 Ml02/gw/hr 




868 Ml02/gw/lir 




1744 /il02/gw/hr 






(Riech et al., i960) 


225 Ml02/gw/hr 


n.s. J 



Temperature Acclimation Temperature and Enzyme Per Cent by 
Oxidative of Activity Per gm Wet Weight which Cold 

Enzymes Measurement or mg Protein Exceeds Warm 

Cold Warm 

Succinic dehydrogenase 
eel liver 
(Precht, 1951) 
goldfish liver 
(Murphy, 1961) 

goldfish muBcle 
(Murphy, 1961) 



15.9 JIl/g^^,/min 
17 /il/g^/min 

8.11 Ml02/gw/™'° 








eel liver 
(Carlsen, 1953) 
eel muscle 
(Carlsen, 1953) 




Malic dehydrogenase 
goldfish liver 
(Murphy, 1961) 

Cytochrome c oxidase 
goldfish liver 
(Murphy, 1961) 

goldfish muscle 
(Murphy, 1961) 


5° 30° 

9.15 ;il02/gw/hr 16.2 

14.7 Ml/gw/™in 22.5 

15U) M l/g/min 
23.4 M l/g^/rain 
.236 M l/mgpr/min 
13.9 Ml02/gw/ni*° 





DPNH cytochrome reductase 
goldfish liver 
(Murphy, 1961) 

TPNH cytochrome reductase 
goldfish liver 
(Murphy, 1961) 

carp gill 
(Eld)erg, 1961) 
eel liver 
(Precht, 1951) 




29 .9 /g^ 










CN inhibition 
goldfish gill 
(Ekberg, 1958) 
goldfish liver 
(EldDerg, 1958) 
goldfish liver 
(Kanungo, Prosser, 1959) 

79.1% Inhlb. 



No significant difference in inhibition by GO, CN, azide. 



Hexose monophosphate 
shunt enzymes 

Acclimation Temperature and Enzyme 

Activity Per gm Wet Weight Per Cent by which 

or mg Protein Cold Exceeds Warm 



Glucose - 6-PO4 dehydrogenase 
Crucian carp gill 
(Ekberg, 1361) 

goldfish liver 
(Murphy, 1961) 

6 - PO4 - gluconic dehydrogenase 
Crucian carp gill 
(Ekberg, 1S61) 

goldfish liver 
(Murphy, 1961) 

Glycolytic enzymes 



44.2 arb. units 




On 298 




11.0 arb. units 









Crucian carp gill 



(Ekberg, 1961) 




Anaerobic acid production 

Crucian carp gill 



(Ekberg, 1961) 

263 (7 hr. day) 

176 (7 hr. day) 


326 (17 hr. day) 

194 (17 hr. day) 


Lactic dehydrogenase 

goldfish Uver 



(Murphy, 1961) 

189 3 /g^ 






lOA inhibition 

goldfish gill 



(Ekberg, 1959) 

52.6% inhib. 

77.4% Inhib. 

Table III. Activity of enzymes from poikilothermic vertebrates acclimated in 
different temperatures. (--Values, inverse acclimation, or Precht's Type 5). (n.s. 
not significant). Enzyme activities in different units of measurement. 



either no effect or a very slight compensation and goldfish liver has 
no such excess of cytochrome oxidase as rat liver (Murphy, 1961). 
Catalase showed no change ( Carassius gill, Ekberg, 1961) or an in- 
verse or t)T)e 5 acclimation (eel liver, Precht, 1951). No differences 
between warm- and cold- acclimated goldfish were found for DPNH 
reductase and TPNH cytochrome reductase (Murphy, 1961). Wide 
variability in P/O ratios led to equivocal results (Kanungo and Pros- 
ser, 1959a; Murphy, 1961). 

The enzymes of the hexose monophosphate shunt show very low 
activity in fish, and their importance is doubtful (Brown, 1960). For 
example, the activity of glucose- 6- phosphate dehydrogenase in gold- 
fish liver is only 3% of that in rat liver (Murphy, 1961). This enzyme 
showed no compensation in crucian carp gill (Ekberg, 1961) and an 
inverse (type 5) acclimation in goldfish liver (Murphy, 1961). Another 
enzyme of the shunt, 6- phospho- gluconic dehydrogenase, showed a 
large compensation in crucian carp gills, but no change in goldfish 
liver homogenates. The suggestion (Kanungo and Prosser, 1959b) 
that there might be increased use of the monophosphate shunt in the 
cold seems invalid. 

The most important metabolic changes in acclimation to cold 
seem to be in glycolytic enzymes. Sluggish fish such as carp are said 
to survive anaerobically in the cold (Blazka, 1958). Active fish such 
as the Kamloops trout show an increase in lactic acid concentration 
in muscle of as much as 4 l/2 times in 9 minutes of exercise and 
elevated lactic acidpersisted for several hours post- exercise (Black 
et al., 1960, 1961). Blood lactate in unexercised trout and salmon is 
high in comparison with mammals and may rise as much as 6 to 10 
fold after exercise (Black et al., 1960). Pyruvate follows the same 
time course as lactate. A trout accumulates lactic acid and pays off 
an C debt (Black et al., 1960); a crucian carp does not accumulate 
lactate but increases its CO excretion (Blazka, 1958). It appears 
that fish rely considerably on glycolytic metabolism. Lactic dehy- 
drogenase activity is high in goldfish liver, and it shows some tem- 
perature compensation (Murphy, 1961). Total acid production by cru- 
cian carp gills was elevated in cold-acclimation, but the CO pro- 
duction was not (Ekberg, 1961). Aldolase was markedly increased in 
carp gills by cold acclimation (Ekberg, 1961). lodoacetate sensitivity 
of goldfish gills was less in the cold. It appears that the most 



striking enzymic increases in fish tissues in the cold are in those of 
glycolysis. However, intermediate acids must ultimately be oxidized 
and the relatively small changes in electron transport enzymes are 
difficult to explain. 

The preceding evidence indicates that some enzymes change and 
others are unaltered in the compensatory acclimation of fish, that 
corresponding enzymes differ for different tissues, and that enzymes 
may either increase or decrease according to temperature. The 
meaning of inverse acclimation (e. g., malic dehydrogenase in gold- 
fish liver and catalase in eel liver) is not clear. In general, the more 
an animal is taken apart, the less is the apparent acclimation. In our 
laboratory Murphy recently examined the activity of numerous en- 
zymes of goldfish liver and has had difficulty obtaining statistically 
significant differences between those from cold- and warm- acclima- 
ted fish. The range of variability is very great for those genetically 
heterogeneous fish and reproducibility of experiments poor. Cer- 
tainly there is no evidence for a general change in activity of all me- 
tabolic systems, and major effects are probably in the integration of 

Non- enzymatic Chemical Changes 

Other changes besides those in enzymes of intermediary meta- 
bolism havebeennotedincold-acclimation. Changes in water content 
may be significant for marine fish. At 1.6 C in sea- water the tide- 
pool fish Girella lost 23% of their water, and they survived only 2 
days, whereas in 45% sea- water at the same temperature no water 
loss was observed and survival was prolonged (Doudoroff, 1938). In 
fresh- water fish, however, an increase of 2% in water content was 
reported for goldfish after 2 days at4 C (Meyer et al., 1956) and a 
decrease after 25 days at 5° C (Hoar and Cottle, 1952). Goldfish 
liver showed no significant difference in water content for 5 C and 
30 C acclimation (Murphy, 1961). 

Protein content of liver from goldfish acclimated at 5 C was 

9.9% and from those at 30° C was 12.5% (Murphy, 1961); no change 

was found in the protein content of muscle. 



Changes in lipids have been reported. Acclimation of goldfish to 
cold was accompanied by increased unsaturation of tissue lipids and 
acclimation to heat by a decreased unsaturation (Hoar and Cottle, 
1952). In cold the tissue phospholipids of goldfish increased in re- 
lation to cholesterol, but no consistent correlation was observed be- 
tween the ratio of cholesterolto phospholipid and thermal resistance. 
Also no good correlation was found between dietary lipid unsaturation 
and thermal resistance, although a high cholesterol diet increased 
resistance to both heat and cold (Irvine etal., 1957). Preliminary ob- 
servations indicate a higher percentage of stearic and palmitic acids 
in liver of 30 C- acclimated than of cold- acclimated goldfish 
(Murphy and Johnston, 1961). Liver of goldfishfrom 5 C had 1.76% 
lipid, from 30 C had 3.97% lipid. The iodine numbers were as fol- 
lows: 30 C, 97.7; 15 C, 100.3; 5 C, 102.3; hence the liver lipid is 
more unsaturated in the cold-acclimated state. Similar changes were 
noted by Hoar and Cottle (1952). In view of the central nervous 
changes to be reported below, it is likely that numerous changes in 
the lipids of cell membranes will be found. 

What mechanisms underlybiochemical changes? Changes in lip- 
ids must depend on differences in some synthetic enzymes. As stated 
above, there is no evidence for involvement of the adrenal cortex and 
conflicting evidence for involvement of the thyroid in temperature 
acclimation of poikilothermic vertebrates. Much more work should 
be done on possible hormonal regulation of metabolism. However, 
present evidence favors a direct effect of temperature. This could 
occur in several possible ways; (1) In the cold, the total metabolism 
is lower than at high acclimation temperatures ; hence metabolic sub- 
strates in general may accumulate, and these may induce more in- 
termediary enzymes at several levels. (2) an intermediate such as 
pyruvate (or lactate) may accumulate because its degradative en- 
zymes have a higher Q than those enzymes forming it, and thus 
this intermediate may reach concentrations which induce an alternate 
path. If B>-C, in the system: 

A ^i^ B ^i^ C 


has a highQ , B accumulates and may induce the enzyme catalyzing 



B (S D. (3) An accumulation of products of one step can repress or 
can stimulate an earlier step in a sequence. The actions (2) and (3) 
may be on enzyme- forming RNA or even on the DNA template. In 
nature the capacityof a particular enzyme- forming system to change 
can form the basis for selection under temperature stress. 

The biological significance of the chemical changes in tempera- 
ture acclimation is uncertain. Many enzymes show enhanced activity, 
some are tmaltered, and a few decreased inaction after cold accli- 
mation. No calorimetric measurements of total energy liberation 
have been made, and determination of P/O ratios for liver mitochon- 
dria have led to equivocal results. The lipid changes are in the dir- 
ection of lower melting points in the cold. A fish or a salamander at 
a low temperature is never as active as at a high temperature. 

Behavior and nervous changes 

In a temperature gradient a fish "selects" a temperature where 
the frequency of spontaneous movements is least; this selection is 
determined by sensory input from cutaneous thermoreceptors and 
is upset by lesions of the forebrain (Sullivan, 1954; Fisher, 1958). 
The "selected" temperature is higher than a low temperature of ac- 
climation and lower than a high acclimation level (Sullivan and Fish- 
er, 1953, 1954), and shifts according to acclimation (Fry and Hart, 
1948a). When maximum swimming speed is measured at different 
temperatures, the optimal temperature rises (Fry and Hart, 1948a), 
and the temperature at which active swimming stops is higher 
(Roots, 1961) as the acclimation temperature rises. 

The O consumption measured in maximum swimming activity 
rises with temperature more rapidly over a low temperature range 
and then more slowly than does the standard or rest metabolism 
(Figure 6). The difference between the two curves (active and stand- 
ard) for fully acclimated fish is considered a measure of extra en- 
ergy available for swimming, the "scope of activity" of Fry. This 
difference curve or scope of activity rises to a maximum in lake 
trout (Salvelinus) at a temperature close to that of maximum cruising 
speed (Gibson and Fry, 1954), and it has been suggested that the 
maximum motor activity is determined by the energy available to 






















/ O 












40 — 

8 16 24 32 40 


Figure 6. Active and standard metabolism and the difference between them when 
measured at the temperatures of acclimation. From Fry and Hart, 1948 (Fig. 5, 
p. 73). 



the fish (Fry, 1947). The change after acclimation of the temperature 
at which swimming is maximum is, according to this view, due to 
compensatory metabolic alterations such as have been described a- 
bove. Certainly no animal can move more rapidly than energy can 
be made available to it. Fisher (1958) has questioned whether the 
maximum metabolism per se determines the cruising speed or 
whether the limit may be imposed in the nervous system. It is possi- 
ble to increase oxygen consumption beyond that at maximum cruis- 
ing speed by electrical stimulation (Basu, 1959). Some fish (trout) 
show two temperature optima for cruising, and these can be altered 
by brain lesions; swimming rate is affected by light intensity (Fish- 
er, 1958). It appears, therefore,that available energy is not the only 
limiting factor for activity. 

In addition to changes in the temperature preferendum and tem- 
perature of maximum cruising speed with acclimation, or the tem- 
perature of sudden reduction in swimming, other measurements in- 
dicate adaptive alterations in the central nervous system. Conduction 
in peripheral nerves isblockedbycold,and the critical temperature 
for cold block declines with cold acclimation (Roots, 1961). Spinal 

reflex movement of the fins of goldfish was blocked at 10 C, 5 C, 
o o o ' o 

and 1 C respectively for fish acclimated to 35 C, 25 Cj and 15 

o o 

C; the reflex persisted at below 1 C in fish acclimated to 5 C. 

Roots has conditioned fish to avoid either a light or dark end of a 

divided aquarium, and also to interrupt their breathing rhythm when 

given a visual stimulus. The cold- blocking temperatures of these 

conditioned reflexes is higher than for simple reflexes, e. g., block 

o o 

occurs at 15 C for 25 C- acclimated fish. Thus a hierarchy of tem- 
perature sensitivity isfound, with midbrain functions most sensitive, 
spinal functions less so, and peripheral nerve least sensitive to cold 
(Roots, 1961). 

Similarly in two species of skate (Raja ) sensitivity of nerve and 
muscle to heat decreases in the following series: myoneural junc- 
tion, nerve conduction, striated muscle contraction, and heart and 
gut muscle activity (Battle, 19 26). It is concluded that important a- 
daptive changes occur in the nervous system during temperature ac- 
climation. Changes in nervous function reflect chemical alterations 
of excitable membranes and subtle changes in interneuronic inter- 
action which are totally unknown. In the absence of the insulative 



changes which occur in homeotherms, the nervous changes under- 
lying behavior are of particular importance in poikilotherms. 

Resistance t o Temperature Extremes 

In nature, the adaptations favoring survival at extremes of heat 

or cold may be more important than compensations in the mid- range. 

Some geographic races of fish (e. g., Salvelinus alpinus) have been 

shown to differ in their lethal temperatures but not in temperatures 

of maximum cruising speed (McCauley, 1958). The literature on 

change of lethal temperature with acclimation is extensive (Brett, 

1956), but very little is known of the responsible cellular changes. 

Some organs are more sensitive than others; for example, the brain 

is more sensitive than the heart, but the chemical bases for such 

differences are unknown. Differences in inactivation temperatures 

have been obseived for some enzymes from thermophilic and meso- 

philic bacteria (Koffler, 1957), and the inactivation temperature for 

amylases, pepsin, and trypsin from fish is lower than for the same 

enzymes from mammals (Chesley, 1934; Vonk, 1941). Acetylcholine 

acetylase of ^ fish (Labrus) brain is maximally active at 25 C and 


is inactivated at 37 G, whereas corresponding temperatures for the 

same enzyme from rabbit brain are 42 C and 47 C respectively 
(Milton, 1958). Changes in lipids as shown by their melting points 
may be important for cell permeability. The effect of endocrines, 
such as the thyroid, on heat death was mentioned above. It may well 
be that more marked changes occur in resistance to temperature 
extremes than as metabolic compensations within the normal range, 
and there may be little relation between the compensations of capa- 
city adaptation and the stress responses of resistance adaptation. 


Acclimation of poikilothermic vertebrates to temperature is 
basically different from that in homeotherms in that compensations 
of poikilotherms tend toward maintenance of relatively constant 



metabolism and behavior with changing body temperature, whereas 
the acclimation of homeotherms tends toward maintenance of con- 
stant body temperature. Some poikilotherms show no compensations, 
but their body processes remain slow (as in hibernation) at low tem- 

No consistent pattern is yet evident for biochemical changes. 
According to in vitro measurements, some enzymes appear to com- 
pensate; many do not. Glycolysis may be most affected in fish. The 
meaning of lipid changes is not clear , although lower melting points 
in the cold are found in fish as well as in peripheral fat of mammals. 

The integrated metabolic system of intact animals shows more 
consistent compensation than isolated enzymes. Undoubtedly hor- 
mones are important in acclimation, as shown by the effects of 
photoperiod. However, there is evidence for direct effects of tem- 
perature, possibly through some sort of enzyme induction. Marked 
differences occur in the response of different organs and tissues to 

Behavioral compensations reflect a heirarchy of differences in 
nervous adaptations with complex conditioned responses being most 
sensitive and pe;ripheral nerve conduction least. These changes in 
sensitivity of nervous systems to cold might be related to alterations 
in membrane lipids. 

Changes in resistance to extreme temperature stress are clear- 
ly indicated by decline in temperatures of both heat and cold death 
with reduced acclimation temperature. Mechanisms of changes in 
resistance to temperature extremes are unknown as are the rela- 
tions between compensation (capacity) acclimation and resistance 




1. Auerbach, M. 1957. Hat die Schilddrvsefurdietemperaturadap- 

tation der Fische eine Bedeutung? Ztsch. f. Fisch. Hilfswiss 
6:60 5-620. 

2. Barrington, E. J. N. and A.J.Matty. 1954. Seasonal variation in 

the thyroid gland of the minnow, Phoxinusphoxinus with some 
observations on the effect of temperature. Proc. Zool. Soc. 
Lend. 124:89-95. 

3. Basu, S. P. 1959. Active respiration offish in relation to ambient 

concentrations of oxygen and CO . J. Fish. Res.Bd. Canada. 

4. Battle, H. I. 1926. Effects of extreme temperatures on muscle 

and nerve tissue in marine fish. Trans. Roc. Soc. Canada Ser. 
m. Sec. V, 20:127-143. 

5. Berg, O., A. Gorbman, and H. Kobayashi. 1959. Thyroid hor- 

mones in invertebrates and lower vertebrates. In Comparative 
Endocrinology, [Ed.], A. Gorbman, (Wiley, New York), pp. 

6. Black, E. C, A. C.Robertson, A. R. Hanslip, and W. Chin. 1960. 

Alterations in glycogen, glucose, and lactate in rainbow and 
Kamloops trout, Salmo gairdneri, following muscular activity. 
J. Fish. Res. Bd. Canada 17:487-500. 

7. Black, E. C., A. C. Robertson, and R.R. Parker. 1961. Some as- 

pects of carbohydrate metabolism in fish. In Compar. Physiol, 
in Heterothermic Animals ,JEd.J, A. W. Martin, (U. Washing- 
ton, Seattle), pp. 89-124. 

8. Blazka, P. 1958. Anaerobic metabolism offish. Physiol. Zool. 




9. Brett, J. R, 1956. Principles of thermal requirements of fish. 

Quart. Rev. Biol. 31:75-87. 

10. Brown, W. D. 1960. Glucose metabolism in carp. J. Cell. Gomp. 

Physiol. 55:81-85. 

11. Brown, W. D. and A. L. Tappel. 1959. Fatty acid oxidation by 

carp liver mitochondria. Arch. Biochem. Biophys. 85:149- 

12. Garlsen, H. 1953. Die cocarboxylasegehalt des Aales (Anguilla 

vulgaris ) . Ztsch. vergl. Physiol. 35:199-208. 

13. Ghesley, L. G. 1934. Influence of temperature upon the amylases 

of cold- and warm-blooded animals. Biol. Bull. 66:330-338. 

14. Dawson, W. R. and G. A. Bartholomew. 19 56 . Relation of oxygen 

consumption to body weight, temperature, and temperature 
acclimation in lizards Uta stansburiana and Sceloporus occi- 
dentalis. Physiol. Zool. 29:40-51. 

15. Doudoroff, P. 1938. Resistance and acclimatization of marine 

fishes to temperature changes. I. Experiments with Girella 
nigrisans. Biol. Bull. 83:219-244. 

16. Ekberg, D. R. 1958. Respiration in tissues of goldfish adapted to 

high and low temperatures. Biol. Bull. 114:308-316. 

17. Ekberg, D. R. 1961. In press. 

18. Fisher, K.G. 1958. An approach to the organ and cellular physio- 

logy of adaptation to temperature in fish and small mammals. 
In Physiol. Adaptation, [Ed.], C. L. Prosser, (Am. Physiol. 
Soc, Washington, D. G.), pp. 3-49. 

19 . Freeman, J. A. 19 50 . Oxygen consumption, brain metabolism and 

respiration movements of goldfish during temperature accli- 
matization; with special references to lowered temperature. 
Biol. Bull. 99:416-424. 



20. Fry, F. E. J. 1947. Effects of the environment on animal activity. 

Publ. Ont. Fish. Res. Lab. (Univ. of Toronto Studies Biol. 
Series 55), 68:1-62. 

21. Fry, F. E. J. and J. S.Hart. 1948a. Cruising speed of goldfish in 

relation to water temperature. J. Fish. Res. Bd. Canada 7: 

22. Fry, F. E.J. and J. S. Hart. 1948b. Relation of temperature to 

oxygen consumption in goldfish. Biol. Bull. 94:66-77. 

23. Fry, F. E. J., J. S. Hart, and K. F. Walker. 1946. Lethal tem- 

perature relations for a sample of young speckled trout, Sal- 
velinus fontinalis. Publ. Ont. Fish. Res. Lab. (Univ. Toronto 
Studies Biol. Ser. 54), 66:1-35 

24. Gelineo, S. 1959. Temperature d' adaptation etconsommationd' 

oxygene chez la scorpene truie Scorpaena porous . Bull. XXV 
Acad. Serb. d. Sci., pp. 37-40. 

25. Gibson, E.S. and F. E.J. Fry. 1954. The performance of the lake 

trout Salvelinus namaycush at various levels of temperature 
and oxygen pressure. Ganad. J. Zool. 32:252-260. 

26. Hannon, J. P. 19 60. Intermediary glucose metabolism in the cold- 

acclimatized rat. Fed. Proc. 19, Suppl. 5:100-105. 

27 . Hart, J. S . 19 57 . Climatic and temperature induced changes in the 

energetics of homeotherms. Rev. Can. de Biol. 16:133-174. 

28. Heroux, O. 1960. Adjustments of the adrenal cortex and thyroid 

during cold acclimation. Fed. Proc. 19, Suppl. 5:82-85. 

29. Hoar, W. S. 1955. Seasonal variations in the resistance of gold- 

fish to temperature. Trans. Roy. Soc. Canada (Series 3), 49; 

30. Hoar, W. S. 1959. Endocrine factors in ecological adaptation of 

fishes. In Comp. Endocrinol., [Ed J , A.Gorbman, (Wiley, New 
York), pp. 1-23. 



31. Hoar, W. S. and M. K. Cottle. 1952. I. Dietary fat and tempera- 

ture tolerance of goldfish. II. Some effects of temperature 
acclimatization on the chemical constitution of goldfish tis- 
sues. Can. J. Zool. 30:41-48, 49-54. 

32. Hoar, W. S. and G. B. Robertson. 19 59. Temperature resistance 

of goldfish maintained under controlled photoperiod. Can. J. 
Zool. 37:419-428. 

33. Irvine, D. G., K. Newman, and W. S. Hoar. 1957. Effects of 

dietary phospholipid and cholesterol on the temperature re- 
sistance of goldfish. Can. J. Zool. 35:691-709. 

34. Jones, S. C, J. G. Phillips, and W.N. Holmes. 1959. Compara- 

tive Physiology of the adrenal cortex. In Comp. Endocrinol., 
[Ed.], A. Gorbman, (Wiley, New York), pp. 582-623. 

35. Kanungo, M. S. and C. L. Prosser. 1959a. Standard and active 

oxygen consumptions of cold- and warm- acclimated goldfish 
at various temperatures. J. Cell. Comp. Phys. 54:259-264. 

36. Kanungo, M. S. and C.L. Prosser. 1959b. Oxygen consumption of 

liver homogenate; oxygen consumption and oxidative phospho- 
rylation of liver mitochondria. J, Cell. Comp. Physiol. 54: 

37. KoffLer, H. 1957. Protoplasmic differences between mesophiles 

and thermophilies. Bacterid. Rev. 21:227-240. 

38. Martin, G. and K. L. A. Tarr. 196l.Phosphoglucomutase, phos- 

phoribomutase andphosphoglucoisomerase of lingcod muscle. 
Can. J. Biochem. Physiol. 39:297-308. 

39. McCauley, R. W. 1958. Thermal reactions of geographic races of 

Salvelinus. Can. J. Zool. 36:655-662. 

40. Mellanby, K. 1940. Temperature coefficients and acclimation. 

Nature 146:165-166. 



41. Meyer, D. K.,B. A. Westfall, and W. S. Platner. 19 56. Water and 

electrolyte balanceof goldfish under conditions of Anoxia, cold 
and inanition. Am. J. Physiol. 184^553-556. 

42. Milton, A. S. 1958. The effect of temperature on the synthesis of 

acetylcholine by the enzyme choline acetylase from mammal- 
ian and fish brain. J. Physiol. 142:25. 

43. Morris, R. W. 1961. Distribution and temperature sensitivity of 

some eastern Pacific cottid fishes. PhysioL Zool. 34:217-227. 

44. Murphy, B. 1961. Tissue metabolism of goldfish (Carassius aura- 

tus) acclimated to warm and cold temperatures. Ph. D. 
Thesis, Univ. of 111. 

45. Murphy, B. and P. Johnston. 1961. Unpublished results. 

46. Olivereau, M. 1955a. Hormone thyreotrope et temperature chez 

la truite arc-en-ciel(Salmogairdnerii) et 1' anguille ( Anguilla 
anguilla). Arch, d' Anat. Mic. e. Morph. Exper. 44:236-264. 

47. Olivereau, M. 19 55b. Temperature et fonctionnement throidien 

chez les Poissons. J. de Physiol. Paris 47:256-258. 

48. Olivereau, M. 1955c. Influence de la temperature sur le fonc- 

tionnement throidien de la truite etudie a 1' aide de 1' iode 
radioactif. Compl. rend. Soc. Biol. 149:536-539. 

49. Precht, H, 1951. DerEinflussder temper aturauf die atmung und 

auf einige Fermente beim all (Anguilla vulgaris). Biol. Zbl, 

50. Precht, H., J. Christophers en, and H.Hensel. 1955.Temperatur 

und Leben. (Springer, Heidelberg). 

51. Precht, H. 1956. Einige Versuche zur temperaturabhangigkeit 

von Lebensprozessen der Hefe Torulopsis kefyr. Biol. Zbl. 



52. Precht, H. 1958. Concepts of the temperature adaptation of un- 

changing reaction systems of cold-blooded animals. In Phys- 
iological Adaptation, [Ed.l, C. L. Prosser, (Am. Physiol. Soc. 
Washington, D. C), pp. 50-78. 

53. Prosser, G. L. 19 58. The nature of physiological adaptation. In 

Physiological Adaptation, [Ed.], C. L. Prosser, (Am. Physiol. 
Soc, Washington, D. C), pp. 167-180. 

54. Prosser, G.L. 1961. Temperature. InComparative Animal Phys- 

iology by G.L. Prosser and F. A. Brown, (Saunders, Philadel- 
phia), pp. 238-284. 

55. Riech, A. , J. A. Belli, and M.E. Blaskovics. 1960. Oxygen con- 

sumption of whole animals and tissues in temperature- accli- 
mated amphibians. Proc. Soc. Exper. Biol. Med. 103:436-439. 

56. Roberts, J. L. 1960. The influence of photoperiod upon thermal 

acclimation by the crucian carp Garassius carassius. Verh. 
Deutsch. Zool. Gesellsch. Bonn, pp. 73-78. 

57. Roberts, J. L. 1961. Preliminary results, by personal communi- 


58. Roots, B. I. 1961. Temperature acclimation in the nervous sys- 

tem of fish: abstract. Fed. Proc. 20:209. 

59. Scherbakov, A. P. 1937. Respiration and temperature adaptation 

in the river lamprey. Bull, de Biol, and Medicin Exper. (U. 
S. S. R.), 3:268-270. 

60. Suhrmann, R. 1955. Weitere Versucheuberdietemperaturadap- 

tation der Karauschen (Garassius vulgaris). Biol. Zutbl. 74: 

61. Sullivan, G. M. 1954. Temperature reception and responses in 

fish. J. Fish. Res. Bd. Ganada 11:153-170. 



62. Sullivan, C. M. and K. C. Fisher. 1953. Seasonal fluctuations in 

the selected temperature of speckled trout, Salvelinus fon- 
tinalis. J. Fish. Res. Bd. Canada 10:187-195. 

63. Sullivan C. M. and K. C. Fisher. 1954. Effects of light on tem- 

perature selection in speckled trout Salvelinus fontinalis . 
Biol. BuU. 107:278-288. 

64. Tashian, R. E. 1957. The relation of oxygen consumption to tem- 

perature in some tropical, temperate and boreal anuran am- 
phibians. Zoologica 42:63-68. 

65. Vernberg, F. J. 1952. The oxygen consumption of two species of 

salamanders at different seasons of the year. Physiol. Zool. 

66. Vonk, H. J. 1941. Die verdauung bei die niederen vertebraten. 

Adv. in Enzymol. 1:371-417. 

67. Wells, N. A. 1935a. Variations in the respiratory metabolism of 

the Pacific killifish Fundulus parvipinnis , due to size, season 
and continued constant temperature. Physiol. Zool. 8:318-336. 

68. Wells, N. A. 1935b. Oxygen consumption by the mud sucker Gil- 

lichthys mirabiUs. Biol. Bull. 69:361-367. 



HUDSON: Dr. Prosser, do you see any signifiance in the 
lower blocking temperature of the peripheral nerves compared 
with the higher parts of the CNS, since functionally, as far as the 
animal is conceirned, if the cord is not responding it would not 
do any good to have the nerves responding? 

PROSSER: I am not sure that I can give you any offhand answer. 
Certainly, the complex behavior which permits feeding and escape 
from predators would be very necessary for survival. Perhaps this 
means merely that integration is the important thing. This cold- 
hardiness of peripheral nerves has been seen before; synaptic 
transmission shows cold block at a higher temperature than nerve 

ADAMS: Dr. Prosser, do you see any change in the lower 
lethal temperature in poikilothermio vertebrates as a result of 
acclimation to higher temperatures, or the converse? One of the 
questions, of course, in homeothermic literature is the inter- 
relationship of cold and heat acclimatization. 

PROSSER: Yes, Precht has reported cases where acclima- 
tion occurred in both directions. But the curves of Fry and his 
associates are quite clear in showing a change in the lower lethal 
temperature with acclimation which may or may not be parallel 
to the change in the high lethal temperature. Both of his curves 
shift in the same direction. 

HART: Do lower lethal and upper lethal temperatures both 
change in the same direction? 

PROSSER: Yes. I wish we knew more about the mechanism 
of this process. I have a feeling that we need to use stress tests. 
We have been looking at changes in the tolerated mid- range of 
temperature. There are virtually no data on the critical tempera- 
tures of enzyme functions. We know very little about changes in 
denaturation temperatures. Dr. Jansky is doing something with this 
and might want to comment on it. 



I should say in respect to the mechanisms of acclimation that 
there has been some indication that one can change the tempera- 
tures of inactivation of enzymes. The prize example of this is in 
the thermophilic bacteria, where Koffler* and others have shown 
that the cytochromes function at temperatures up to 70 C, where- 
as the corresponding proteins from mesophils are knocked out at 
35 G. This is a fantastic difference. It must mean that there is a 
difference in tertiary structure in the same enzyme protein. 

HART: I wanted to ask you aboutthose curves which you showed 
of activity metabolism versus temperature — was it for the green 
sunfish, which is a different species, or was it for the goldfish you 

PROSSER; The activity curve that I showed you was for the 
green sunfish. Your curves have been for goldfish and they were 
smaller goldfish than we used. We have not been able to get such 
complete curves for the goldfish; that was the reason I did not 
show you goldfish data. We have some curves, but for some rea- 
son we have not had as good luck getting complete swimming 
curves for them as for the green sunfish. 

HART: Those are beautiful curves. These curves agree with 
the general concept that Fry developed, which is that the activity 
would be determined by the difference between standard and active 

PROSSER: Yes, I think that is so. 

HART: I wondered if you had any data on the resting versus 
active metabolism to compare with those active metabolism data? 

PROSSER: Not for the green sunfish. 

HART: Does this conflict with Fry's concept? 

♦Koffler, H. 1957. Enzyme of thermal bacteria. Bacteriol. Rev. 21:227-240. 



PROSSER; No, I do not think it conflicts. All I am saying is 
that I think that Fisher's data* suggests that there are central 
nervous components which are involved in the swimming responses 
in addition to the metabolic ones. The difference between our 
curves and your data** is that your activity curves rise to a peak 
and then drop off rather gradually. The ones which Dr. Root has 
obtained come up to a peak as you saw, and drop off very steeply. 

HART: Yes, but the active metabolism may drop off very 
rapidly, too. 

PROSSER; Yes, I think it does. In the data which I showed 
you from Kanungo, the curves would come up to a maximum. The 
shape is somewhat different from those of Fry . The metabolism 
of these goldfish was measured at different temperatures. It was 
not measured at the temperature of acclimation only as it was in 
Fry's data. 

HANNON; In the data that you have just presented, I have seen 
a number of instances where the effects of temperature on poikilo- 
therms and homeotherms are quite similar. For example, in many 
poikilotherms acclimatization to cold is accompanied by an increased 
metabolic rate. We see this same effect in small mammals such 
as the rat. 

PROSSER: For a different reason, though. 

*Fisher, K. C. 1959. Adaptation to temperature in fish and small mammals. 
Physiological Adaptation, pp 3-49. Ed. C. L. Prosser. Amer. Physiol. Soc., 

**Fry, F. E. J, and J. S. Hart. 1949. Cruising speed of goldfish in relation to 
temperature. Jour. Fish. Res. Bd. Canada, 7:169-175. 

AFry, F. E. J. and J. S. Hart. 1948. Relation of temperature to oxygen con- 
sumption in goldfish. Biol. Bull. 94:66-77. 



HANNON; This may or may not be true; I do not feel that we 
have enough evidence at the present time to justify any firm con- 
clusions either way. One basic difference between these two t3T)es of 
animals that should be noted, however, is the decline in the meta- 
bolic rate of the poikilotherm when he is exposed to high ambient 
temperatures. This is quite in contrast to the response of the homeo- 
therm, in which exposure to high ambient temperatures has either 
no effect on or increases the metabolic rate. 

This high temperature decline in the metabolic rate in fish and 
other poikilotherms is most interesting to me from the standpoint 
of its similarity to the effect of temperature on many enzymes. 
There are a great many enzymes that show increasing activities 
with increasing temperature until some critical point, or tempera- 
ture optimum, is reached. Beyond this temperature optimum the 
activity declines. In many enzymes this decline at high temperatures 
is reversible, provided the point of protein denaturation is not 
reached. 1 would imagine that a similar reversible inactivation 
would also apply to the overall respiratory metabolism of poikilo- 

PROSSER; Yes, but I do not think that the same thing is hap- 
pening here. This is a result of acclimation. The direct metabolism 
temperature effect is what one finds in short term exposure to the 
heat. The acclimatory reduction in metabolism at high temperature 
takes days to develop, just as does the increase in metabolism in 

HANNON: To the best of my knowledge, instead of reducing 
metabolism at high temperatures, as the poikilotherms do, the 
mammals increase their heat loss. Thislfeelis a basic difference. 

At the cellular and sub-cellular level, I was quite impressed 
by the many striking similarities between the metabolic activities 
of poikilotherms and homeotherms following cold-acclimatization. 
In fact, I do not feel there are as many incongruities here as you 
do. For instance, you have given a number of examples where cold 
acclimatization leads to an increase in the metabolic rate of the 
intact animal. You have also shown with whole cell preparations 



tiiat this increase is reflected in a similar increase in the meta- 
bolic activity of several tissues such as liver and muscle. We find 
essentially the same results with small mammals. Then to carry 
such studies a step further you have also given a number of instances 
where cold acclimatization in the poikilotherm produces changes 
in enzyme activity that are the same as we find in rats. I would 
include among these latter effects succinic dehydrogenase, cyto- 
chrome oxidase, glucose- 6- phosphate dehydrogenase, and lactic 
dehydrogenase as well as DPNH and TPNH cytochrome c reductase 
which are not affected by prolonged cold exposure in either type 
of animal. It is true that there are several instances where poikilo- 
therms and homeotherms are at first glance quite different in 
their responses to prolonged cold exposure. This, however, may 
or may not be significant since even in one species cold exposure 
can lead to quite a variety of effects. 

There are a number of factors that can influence the nature 
of the results obtained from in vitro tissue metabolism studies. 
In intact cell preparations, for example, we have the problem of 
quite limited exogenous substrate permeability or utilization. This 
is particularly true for those substrates that exist in an ionic form. 
But it is also true for such a common metabolite as glucose. In 
vitro metabolic rate of whole cell preparations, therefore, is 
largely dependent upon the availability of endogenous substrate, 
and we must be quite cautious in interpreting them. 

PROSSER: That is why we used homogenation. 

HANNON; Homogenates are also notoriously bad for oxidizing 
free glucose. They will not do it like the intact animal will. 

PROSSER: We have used succinate, too. 

HANNON: Practically everybody, I think, has reported an 
increase in succinate oxidation in the Liver and muscle of cold- 
acclimated mammals. Skin has also shown this increase. 

PROSSER: Do you find an increase in the monophosphate 
shunt enzymes ? 



HANNON: In our own work we have found a decrease in both 
liver and muscle after one month of acclimatization. Other people 
who have acclimatized their animals for a much longer period 
find no change in the system. This brings up another question: A 
number of investigators, includingsome of the workers at Dr. Hart's 
laboratory, Heroux in particular, have found that the liver metabol- 
ism of animals that are subjected to seasonal, outdoor acclimatiza- 
tion is the same in both summer and winter. Along similar lines, 
we have found that liver metabolism varies with the duration of expo- 
sure. It goes through a peak— in our particular circumstances at 
one month— and then it falls back to the normal levels. It would 
appear then, that the initial response to alow temperature, at least 
in this organ, is an increase in metabolism per unit mass of tissue. 
With longer exposure, however, we find an increase in the rela- 
tive size of the liver. Metabolic ally, this increase in mass replaces 
the increase in unit activity, and theliverthus retains a high meta- 
bolic rate by virtue of its size.Inotice in your data that practically 
all of the metabolic rates are expressed as oxygen consumption per 
gram of tissue, and there is no indication of whether the relative 
mass of tissue has changed. You do have evidence that the protein 
content does change, but I would like to ask whether there were any 
changes in relative liver mass comparable to those we have observed 
in rats. 

PROSSER: Dr. Murphy has found the changes in liver size and 
in the same direction that you find them. That is, the liver of a 
goldfish that has been held at 30 C is very small, whereas the one 
that has been held at 5 C is large. We thought it has more fat on 
a unit weight basis. It does not. It has less fat. You are finding 
that the fatty acid metabolism increases, are you not? 

HANNON: Yes, but we have studiedonly the liver. Dr. Depocas, 
I believe, has found that the intact animal can oxidize fatty acid 
at greater rates in the cold. Am I correct, Dr. Hart? 

HART: Yes, but not associated with acclimation. There is a 
greater elevation of oxidation in the cold, but there was no change 
associated with acclimation. 



PROSSER: What about the Krebs cycle enzymes? 

HART: No alteration. 

PROSSER: How about the glycolytic ones? 

HANNON: In general, we have found that the overall metabolic 
capacity of the Krebs cycle is increased by pro longed cold exposure. 
In the one month cold-exposed animal this is evidenced by an in- 
crease in the activity of all the Krebs cycle oxidases we have stu- 
died. At the enzyme level, cold exposure may or may not lead to 
an increased activity. Thus, in the electron transport system of the 
liver, it was found that the activities succinic and malic dehydro- 
genase and cytochrome oxidase were elevated following one month 
in the cold. DPNH-cytochrome reductase, on the other hand was 
unaffected. Similar increases in Krebs cycle activity have also been 
seen in muscle. On the basis of these data, we might tentatively con- 
clude that the primary effect of cold exposure (at least after one 
month) is an increase in the metabolic capacity of those reactions 
that are rate- limiting, e.g. succinic and malic dehydrogenase and 
cytochrome oxidase. This conclusion, however, may have tobe modi- 
fied for animals that have been exposed for intervals longer than 
one month. Also, we have only limited data on how cold exposure 
affects muscle tissue, and we have no idea how changes in relative 
mass might affect the results that are obtained. 

PROSSER: Were these Krebs cycle activity measurements cal- 
culated on a unit weight basis? 

HANNON: That is correct. The increase in oxidation will pro- 
bably disappear with exposures that would lead to an increase in the 
relative amount of tissue. In response to an earlier comment, I 
should mention that fatty acid oxidation, at least the oxidation of 
palmitic acid, does proceed at a greater rate in the liver tissue cf 
the cold-acclimatized rat. 

PROSSER: The factor of exposure time, I am sure, is very 
important. We have used periods of one to three weeks because 
this agrees with acclimation time for lethal temperatures. How- 
ever, Dr. Murphy showed me some data taken from experiments 



on muscle where both glucose and succinate were used as sub- 
strates. Her values for cold- acclimated fish seemed not to change 
at all. They stayed high, so that after some weeks the value in the 
cold was higher than the warm, hence the acclimation response 
was reduced activity in the warm acclimated animals with virtually 
no change in the cold acclimated animals. All the enzyme data that 
I gave you were from fish that had been on a 12 hour photoperiod. 

HANNON: A number of things are variables here that we know 
very little about. And one of them is the variable of intermittent 
exposure. Other factors are the effects of changes in light as well 
as changes in the age or changes in the size of the animal. All of 
these variables, at least potentially, could lead to a big difference 
in the type of response you get. 

HART: I would like to ask one other thing in connection with 
the enzyme work; since there appear to be large changes asso- 
ciated with the overall activity of the animal during acclimation, 
I wonder if enzymes associated with the maximum metabolism might 
be worth investigating. 

PROSSER: How are you going to find these? 

HART: I wonder if the cytochrome oxidase activity would have 
some bearing on this. 

JANSKY: We could expect some differences in the cytochrome- 
oxidase activity and especially in the shifting of the optimum of 
this enzyme according to the temperature of acclimation. We have 
some evidence about it on insects. 

HART: What would be your opinion of this approach, Dr. 

JANSKY: I would say we could find some differences in the 
maximum metabolism, and especially in the shifting of the tem- 
perature for maximal enzyme activity. 



PROSSER'. Yes, I think the shifting of the optimum is very 
important, and this is one of the things that we are proposing to 
do soon, but I am not sure that I would agree that the cytochrome 
oxidase system is necessarily the .limiting one for activity meta- 
bolism since these enzymes seem always present in excess. Are 
you implying that this is the principal route for activity as opposed 
to standard metabolism? 

HART: I would not like to say anything about that now, since 
we will hear more evidence about this later on. 

HANNON: I think at this time we might point out that after 
one month of acclimatization you do find an increase in cytochrome 
oxidase activity. I feel that it is important to keep in mind that 
practically all hydrogen transport from Krebs cycle oxidations 
eventually channels through this particular enzyme. Thus, it would 
seem likely that cytochrome oxidase may not be as much in excess 
as the activity measurements might suggest. In fact, it may even 
be rate limiting. If this proves true then cytochrome oxidase would 
be a very good index of maximal metabolic capacity. 

PROSSER: Yes, but we found very little effect on any of the 
Krebs cycle enzymes that we have looked at. I do not understand 
the inverse acclimation of some of them. I doubted the phenomenon 
on the basis of Precht's experiments. However, we came up with 
two enzymes which show it, and it is highly significant. 

IRVING: One of the things that impressed me is that when we 
look at the changes of the influences of temperature on various 
functions we expect to see some more or less continuous slopes, 
that is, something which will relate the rate to temperature in the 
form of a curve; and yet many changes of behavior occur explo- 
sively at given temperatures, whether it be the flight or the biting 
of the insects, or the sensation of same. Insects do not half fly. 
They either completely fly or they are completely quiescent. Of 
course, they also have certain reverse or discontinuous changes — 
as, for example, in the discharge of cold receptors, which apparently 



constitute a whole area or population with different temperature 
thresholds. Abrupt physiological changes in temperature constitute 
the animal's own analysis of what the situation is because he must 
then expunge all other influences of everything else at those par- 
ticular moments. 

PROSSER: This is something we are going to explore. We hope 
soon to probe in the brain with electrodes and see if we can find 
some different recording of electrical activity at different tempera- 



William R. Dawson* 

The possession of homeothermy by birds and mammals has ex- 
ercised a major influence on their evolution, both through the bio- 
logical opportunities it has afforded and through the physiological de- 
mands it has imposed. The evolutionof the mechanisms responsible 
for this condition merits consideration not only because of its im- 
portance to these groups of vertebrates, but also because it com- 
prises a major step in a general trend within the Animal Kingdom 
toward increasing control of internal state. The present discussion 
will deal primarily with the evolution of the mechanisms of tempera- 
ture regulation in birds, although reference will be made to mam- 
mals where comparisons are appropriate. The development of tem- 
perature regulation in this latter group is treated in detail elsewhere 
(Johansen, 1962). 


Current concepts of the origin and early deployment of birds are 
largely a matter ofdeduction, owing to the very incomplete fossil re- 
cord. The structure of the earliest known bird. Archeopteryx litho - 
graphic a . from the upper Jurassic of Bavaria, places the origin of 
the class among the thecodont reptiles (Swinton, 1960). Birds appear 
to have arisen from a single line which appeared with the radiation 
of this reptilian order in the Triassic. The stage in the development 
of this line at which homeothermy was achieved is unknown, and for 
this reason subsequent references to the establishment of this condi- 
tion in the "avian evolutionary line" are intentionally vague. Swinton 
(1960) suggests that the immediate antecedents of birds were arbore- 
al and at least partially homeothermic, and that true flight was not 

♦Preparation of this paper was supported in part by a grant from the National 
Science Foundation (G-9238). 



achieved until after the appearance of effective temperature regula- 

The adaptive radiation of birds, in good part made possible by 
the possession of homethermy, apparently began shortly after birds 
first appeared. However, it did not proceed at a constant rate. As 
far as can be determined from known fossils, the major flowering 
of avian evolution occurred early in the Tertiary. By the end of the 
Eocene most of the known orders of birds had appeared, and by the 
end of the Miocene, most Recent families of birds were probably in 
existence. Today thereareover 8,000 species of living birds, repre- 
senting some 28 orders and 161 families. Six orders and 39 families, 
not counting fossils of uncertain taxonomic position, are known to 
have become extinct (Storer, 1960). All contemporary birds appear 
highly modified for their respective adaptive niches, and none is es- 
pecially primitive. Beddard (1898:160-161) concluded, "the few spe- 
cially reptilian features in the organisation of birds have, so to speak 
been distributed with such exceeding fairness through the class that 
no type has any great advantage over its fellows;." In contrast, 
mammals include both primitive and highly advanced types. Among 
the former, the monotremes, though specialized in some respects, 
have many of the structural features of therapsid reptiles (Simpson, 


Central body temperatures of active birds are generally main- 
tained between 38 C and 43 C, with the limits for individual species 
being narrower (King and Earner, 1960). Perhaps utilization of this 
band of temperatures resulted from a compromise between two un- 
favorable ranges of temperature (Burton and Edholm, 1955). On one 
hand, it was far enough above the rather moderate temperatures 
which apparently prevailed in the Triassic and Jurassic (Brooks, 
1949) so thatphysiological changes required to cope with minor fluc- 
tuations in ambient temperature would be relatively small. On the 



other hand, it was sufficiently far below the lethal temperature level 
(about 47 C for contemporary birds) that moderate elevations of 
body temperature resultingfrom activity, for example, could be sus- 
tained without injury. 

The fact that the central body temperatures of birds do fall in 
a fairly narrow band indicates that relatively little diversification 
of this physiological character has occurred in the evolution of birds 
subsequent to the establishment of homeothermy. One consequence 
of this conservatism has been to render the fundamental level of 
body temperature non- adaptive to climate (Scholander etal., 1950a; 
Irving and Krog, 1954; Scholander, 1955; and Irving, 1960), although 
temporary hjqjothermia and hyperthermia appear to have roles in 
short-term adjustments to cold and heat, respectively, in some spe- 
cies. Steen (1958) found that freshly captured small birds adjusted 
to winter conditions in Oslo, Norway. These included Titmice ( Parus 
major). Green Finches ( Chloris chloris ). Bramblings ( Fringilla 
montifringilla ) , House Sparrows ( Passer domesticus ) , Tree Spar- 
rows ( P. montanus) . and Redpolls ( Acanthis flammea ) . They became 
hypothermic by as much as 9 to lo C when exposed to cold at night. 
However, hypothermia did not develop in these birds when they were 
experimentally acclimated to -10 C. Bartholomew and Dawson 
(1958) regard hyperthermia as a regular feature of the response of 
birds to heat. The tolerance by these animals of as much as 4 C in 
excess of normal levels allows establishment of a favorable condi- 
tion for heat transfer from body to environment when environmental 
temperatures rise to near the level of body temperatures maintained 
in cool environments. This response is of great significance in arid 
regions because it reduces the demands for evaporative cooling from 
what they would be in hot weather if body temperatures were main- 
tained constant. The statement concerning the non-adaptiveness of 
body temperature of course pertains to central body temperatures 
and not to the temperatures of the peripheral tissues, particularly 
in the legs, of birds. Variation in the temperatures of these tissues 
comprises an important component of physical thermoregulation in 
these animals (Irving and Krog, 1955). 

The fact that the general level of body temperature adopted by 
birds exceeds that of mammals may confer a slight advantage in 
warm environments, but its effect on heat exchange is probably in- 
consequential in cold ones. The difference in thermal levels for the 



two classes probably reflects nothing more than differences in the 
temperature relations of the independent reptilian stocks from which 
they emanated. Variations in levels of activity and lethal body tem- 
peratures comparable in extent to the differences separating birds 
and mammals can be found among contemporary reptiles, desert 
lizards and snakes for example, (Cole, 1943; and Cowles and Bo- 
gert, 1944). 

The stabilization of body temperature at a high level in birds 
may have demanded physiological adjustments beyond those con- 
cerned with the establishment of thermoregulatory capacities, even 
if the antecedents of the first homeotherms in the avian line had 
utilized high body temperatures for activity in the manner of many 
contemporary reptiles, particularly lizards. These animals, despite 
their utilizing body temperatures similar to those of homeotherms 
for activity (see Cowles and Bogert, 1944; Norris, 1953; and Fitch, 
1956), apparently have not developed the capacity for prolonged 
existence at a high thermal level. Wilhoft (19 58) found that main- 
tenance of fence lizards (Sceloporus occidentalis ) at their activity 
temperature (34 C) for approximately three months resulted in the 
death of some animals, increased frequency of molting in some, and 
increased thyroid activity in all. None of these changes was observed 
in the control animals, which were allowed a more normal thermal 
pattern in which warm body temperatures alternated with cooler 
ones. The duration of the daily periods spent at warm body tem- 
peratures by heliothermic lizards such as Sceloporus and Uma is 
apparently controlled in partby the parietal eye (Stebbins and Eakin, 
19 58). Elimination of this structure or shielding it from radiation 
significantly increased the extentto which the lizards exposed them- 
selves to sunlight. 


The evolution of the complex array of processes on which home- 
other my depends must have involved many steps. It has been possi- 
ble to gain some insight into the probable nature and sequence of 



these steps in mammals bycomparisonof species representing pri- 
mitive and more advanced levels (see Eisentraut, I960) in develop- 
ment of homeothermy. This approach was employed in the classic 
study by Martin (1903) and has most recently been utilized by Johan- 
sen (1961). Such an approach is less useful in attempting to trace the 
evolution of temperature regulation in birds because of the absence 
in the contemporary avifauna of especially primitive forms with re- 
spect to attainment of homeothermy (see ''The Historical Back- 
ground"). In the subsequentdiscussion a good deal of dependence has 
been placed on information concerning the ontogeny of temperature 
regulation in birds and on data on the physiological responses of con- 
temporary reptiles to temperature. Due regard has been given the 
difficulties of deriving evolutionary interpretations from such in- 
formation: Ontogeny may only recapitulate phylogeny when expedi- 
ent, and contemporary reptiles are for the most part far removed 
from any line of direct importance to the history of birds. 

Behavior of Significance in Temperature Regulation 

Significant behavioral patterns in management of temperature 
relations are widespread and presumably of considerable antiquity 
among vertebrates. The ability of fishes to select particular tem- 
peratures in experimental gradients is well known, and it appears 
that this type of behavior plays a role in the distribution of at least 
some species in nature (Sullivan, 1954). More pertinent to a con- 
sideration of the evolution of temperature regulation in birds is 
the abilityof reptiles under favorable conditions to control their body 
temperatures by behavioral means (Cowles and Bogert, 1944 ; Bogert, 
1949a, 1949b; Norris, 1953; Fitch, 19 56; and Saint- Girons and Saint- 
Girons, 1956). Selection of suitable microclimates and absorption of 
solar radiation allow many species to establish characteristic and, 
in some cases, very high levels of body temperature when they are 
abroad and active. The extent of the control of body temperature 
which can be achieved by behavioral means when sufficient solar 
radiation is available is indicated by the Andean lizard ( Liolaemus 
multiformis) , which Pearson (1954) found could maintain a tempera- 
ture above 30 C by basking, even though nearby shade temperatures 
were at or below freezing. 



Undoubtedly the first homeotherms in the avian line received a 
considerable legacy of thermally significant behavioral patterns 
from their reptilian antecedents. Indeed, the evolution of physiologi- 
cal mechanisms for regulation of body temperature may well have 
been originally concerned with augmenting thermoregulatory be- 
havior. As physiological capacities for temperature regulation im- 
proved, behavior came to occupy the ancillary role in the manage- 
ment of the heat economy evident in birds today. Many species, as 
a result of their migratory habits, are able to exploit various en- 
vironments on a seasonal basis and to evade unfavorable conditions 
to a great extent. Birds resident in hot climates modify the impact 
of their environments to some extent by utilizing shade, minimizing 
activity, and, in some cases, bathing during the heat of the day (Daw- 
son, 1954). In a few instances birds resident in cold climates also 
employ behavioral mechanisms incopingwithwinter conditions. For 
example, at night ptarmigan (Lagogus) utilize the shelter afforded by 
burrows in the snow (Irving, 1960) and Creepers ( Certhia brachyda - 
tyla) huddle together in bunches of 10 to 20 (Lohrl, 19 55). In general 
it appears that birds, particularly the smaller ones, are less suc- 
cessful in evading the extremeconditionsof their environments than 
their mammalian counterparts. The fact that most desert birds are 
diurnal and fail to take advantage of the shelter afforded by under- 
ground burrows forces them to contend with heat as well as aridity. 
This has an important effect on their water economies because it 
requires rapid rates of evaporative water loss (Bartholomew and 
Dawson, 1953; Bartholomew and Cade, 1956; Dawson, 1958). Such 
is not the case in most small desert mammals, which are fossorial 
and nocturnal (see Schmidt- Nielsen and Schmidt- Nielsen, 19 52). 
The failure of birds in cold climates to utilize underground bur- 
rows and nests also deprives them of effective means of protec- 
tion utilized by many mammals. 

With the establishment of homeothermy in the avian line, the 
general thermal requirements for development became restricted 
to a fairly narrow range of temperatures a few degrees below the 
level of body temperature in adults. This restriction must have been 
accompanied by the evolution of elaborate patterns of parental be- 
havior, which are evident in contemporary birds (Kendeigh, 1952). 



Incubation in birds is nicely regulated so that the eggs are maintain- 
ed within the appropriate temperature range most of the time, de- 
spite external conditions (compare Huggins, 1941; Irving and Krog, 
1956; and Eklund and Charlton, 1958). Incubation is facilitated in 
many birds by the development under hormonal control of a well- 
vascularized and defeathered incubation patch (Bailey, 1952). The 
uniformity of incubation temperatures for most species indicates 
that little diversification of the thermal requirements for develop- 
ment occurred after they were originally defined. 

Considerable diversity of parental behavior with respect to the 
post-hatching phase of development in birds is evident, and this is 
consistent with the wide variation in the state of maturity of the 
young on emerging from the egg (see "Patterns in the Ontogeny of 
Temperature Regulation"). The behavior of the parents nicely com- 
pensates for any thermoregulatory deficiencies in the young, so that 
development proceeds under essentially homeothermic conditions 
(Kendeigh, 1952) independent of external temperatures. The activi- 
ties of the parent birds include not only protecting the young from 
cold by brooding, but also shielding them from solar radiation under 
certain conditions, as noted in pelicans ( Pelecanus erythrorhynchos 
and P. californicus ) and Herons ( Ardea herodias ) by Bartholomew 
et al. (1953) and Bartholomew and Dawson (19 54a) and in Nighthawks 
(Chordeiles minor) by Howell (1959). As young birds attain effective 
temperature regulation, the role of parental behavior in their heat 
economy progressively declines. 

Thermostatic Mechanisms 

It is difficult to trace the origin of the neural mechanisms con- 
trolling temperature regulation in birds, if only because these mech- 
anisms have thus far been characterized in only the most general 
terms. Regulatory activity appears to be controlled principally by 
thalamic or hypothalamic centers (Rogers, 1928; Rogers and Lackey, 
1923), although some activity persists after these centers have been 
eliminated (Kayser, 1929a, 1929b). In the Domestic Fowl ( Gallus 
gallus) shivering can beelicitedby stimulation of cutaneous cold re- 
ceptors or of central areas through reduction of skin temperature or 
central body temperature, respectively (Randell, 1943), Panting 



appears to be controlled by a center in the midbrain, judging by von 
Saalfeld's observations on Rock Doves ( Colum ba livia) , and cannot 
be elicited by peripheral stimulation (Randall, 1943). Pan ting is un- 
affected byvagotomy inthe rockdove.butis abolished by this opera- 
tion in the Domestic Fowl (Hiestand and Randall, 1942). 

Whatever the details of the original and present features of the 
mechanisms governing temperature regulation in birds, it is appar- 
ent that they must have been dependent fundamentally on a capacity 
for the detection of absolute temperature (as opposed to detection 
of temperature change) . This capacity is not an original develop- 
ment by homeotherms, but also must be present in many poikilo- 
ther ms , judging by the widespread distribution of temperature selec- 
tion among them (Fry, 1958). The functional basis of absolute tem- 
perature detection is little understood, although analysis of the non- 
adapting fraction of the thermal sensitivity of some peripheral re- 
ceptors is providing some information (Bullock, 1955). The means by 
which it is accomplished in the behavioral regulation of body tem- 
perature by reptiles is unknown, but recent work (DeWitt, personal 
communication) suggests that in the lizard ( Dipsosaurus dorsalis) . 
and presumably in other species, it is actually the temperature of the 
brain or one of its parts thatis regiilated. Rodbard (1948) claims to 
have demonstrated the existence of a thermally sensitive area in the 
hypothalamus of the turtle, which controls blood pressure, and on 
this basis suggests that the thermoregulatory centers of homeo- 
therms evolved from a hypothalamic area controlling circulatory ac- 
tivity. Such a conclusion seems premature considering the absence 
of information on the neural mechanisms responsible for controlling 
thermoregulatory behavior and panting in reptiles. 

Metabolic Level and Chemical Regulation 

The basal metabolic rates of birds and mammals are as much as 
eightfold greater than the resting metabolic rates of reptiles of com- 
parable size at the same body temperature (Martin, 1903; Benedict, 
1932, 1938; and Daws on and Bartholomew, 19 58), and the intensifica- 
tion of metabolism has apparently comprised a most important step 
in the evolution of homeothermy. Zeuthen (19 53) has suggested that 
this intensification was achieved through prolongation of the develop- 
mental phase in which metabolism and size are nearly proportional. 



Hemmingsen (1960) emphasizes that the transition from a poikilo- 
thermic to a homeothermic metabolic level was to a great extent de- 
pendent on an increase in the area of the respiratory surfaces. Ob- 
viously this transition exercised a profound effect on the respiratory, 
circulatory, and other organ systems of the nascent homeotherms. 
The bolsteringof the capacity of these systems, which served to sus- 
tain heightened demands of metabolism, probably contributed subse- 
quently to the development of the thermoregulatory processes. For 
example, modifications of the cardiovascular system, which allowed 
operation with a higher cardiac output and higher systemic blood 
pressure, probably improved capacities for transport of heat over 
those possessed by reptiles. 

The elevation of the general level of metabolism made possible 
the development of effective chemical regulation. Such regulation 
appears to provide the initial means by which young birds control 
body temperature in moderate to cool environments. For example, 
the development of temperature regulation in young House Wrens 
(Troglodytes aedon )at an ambient temperature of 18 C is closely 
correlated with the appearance of muscle tremors (Odum, 1942). 
Similarly, in young domestic fowl, the ability to maintain body tem- 
perature at a high level during exposure to an ambient temperature 
of 20 C initially appears to be associated with the acquisition of the 
ability to shiver (Randall, 1943). These observations suggestthat the 
development of chemical thermoregulation was one of the initial 
steps in the evolutionofhomeother my in birds. Martin (1903) reach- 
ed a similar conclusion for the evolution of this condition in mam- 
mals on the basis of his studies of temperature regulation in adult 
monotremes, marsupials, and placentals. If this suggestion is cor- 
rect, the advent of chemical thermoregulation must have provided 
the initial means by which a level of body temperature established 
under favorable conditions as a result of suitable behavioral pat- 
terns and of an intensified level of metabolism could be maintained 
in cooler surroundings. 

The principal development in the evolution of chemical thermo- 
regulation in the avian line has concerned mechanisms for varying 
muscular heat production. Increasing muscle tonus and, ultimately, 
shivering are the principal means besides activity by which con- 
temporary birds augment their heat production (Steen and Enger, 
19 57; King and Farner, 1960). The ability to sustain elevated levels 



of heat production for long periods of time appears well developed 
in many birds, particularly small northern species. These birds, 
which include the Snow Bunting (Plectrophenax nivalis ) studied by 
Scholander et al. (1950b), the Yellow Bunting ( Emberiza citrinella ) 
studied by Wallgren (1954) , the several previously mentioned species 
studied by Steen (1958), the Evening Grosbeak( Hesperiphonavesper - 
tina), and Red and White- winged Crossbills (Loxia curvirostris ) and 
L. leucoptera ) studiedby Dawson and Tordoff( 19 59 and unpublished) , 
have lower critical temperatures well above the ambient tempera- 
tures which they encounter in their habitats during winter. Although 
it would seem advantageous for these animals to be able to supplant 
the thermogenesis achievedby shivering with that stimulated by hor- 
monal substances in meeting their requirements for elevated heat 
production, they appear not to possess the latter mechanism (Hart, 

Once the intensification of metabolism had been achieved in the 
avian line, relatively littlediversificationof metabolic level appears 
to have occurred, other than that associated with the diversification 
of body size. Although the relation of basal metabolism to body 
weight in birds is less well known, particularly at the extremes of 
size, and apparently more complex than that for mammals (King and 
Earner, 1960), it appears similar in arctic, temperate, and tropical 
species. This has led Scholander, Irving, and associates to empha- 
size that basal metabolic rate is fundamentally no n- adaptive to cli- 
mate (Scholander etal., 1950a; Scholander, 1955; Irving et al., 1955, 
and Irving, 1960). 

Physical Regulation 

The various components of physical regulation, which serve to 
alter heat lossby modification of rates of heat transfer and evapora- 
tion, probably did not arise simultaneously in the avian evolutionary 
line. The ability of contemporary reptiles such as the lizards Dipso- 
saurus dorsalis and Sauromalus obesus to pant when heated (Cowles 
and Bogert, 1944; Dawson and Bartholomew, 1958; Dill, 1938) sug- 
gests that this mechanism for enhancing evaporative cooling could 



have appeared in this line as a legacy from its reptilian antecedents. 
The early development of panting in young birds, e. g., albatrosses 
(D iomedea immutabilis and D. nigripes ) , studied by Bartholomew 
and Howell (1961); herons, studied by Bartholomew and Dawson, 
(1954a); and House Wrens, studied by Kendeigh (1939), likewise sug- 
gests that this mechanism is of considerable antiquity in birds. If 
panting was inherited from the reptilian antecedents of birds, its 
function was apparently bolstered by subsequentchangesinthe car- 
diovascular and respiratory systems, associated with the intensifi- 
cation of metabolism. Rates of evaporative water loss by panting 
lizards are a fifth or less those of panting birds of comparable size 
at the same body temperature (compare data on the birds Pipilo 
fuse us , P . aberti , and Richmondena cardinalis (Dawson, 19 54, 1958) 
and on the lizard Dipsosaurus dorsalis (Templeton, i960)). It would 
be of considerable value to an understanding of the origin of panting 
to determine whether or not the midbrain center controlling panting 
in the Rock Dove (von Saalfeld, 193 6), and presumably in other birds 
as well, is homologous to the neural apparatus governing this ac- 
tivity in reptiles. 

The apparently universal distribution of panting among birds 
contrasts to the situation in mammals, which may bring about eva- 
porative cooling at high temperatures by panting, sweating, or be- 
havioral means. It is probably significant that panting, though it is 
not a highly effective means of heat dissipation, is the only one of 
these mechanisms which would not interfere with flight by marring 
the integrity of the plumage of birds. Flight has undoubtedly created 
special problems for these animals because of the high level of heat 
production which it involves. Dissipation of this heat must require 
extensive evaporative cooling, owing to the effectiveness of the in- 
sulation provided by feathers. The development of the avian res- 
piratory system was probably influenced by this need for evapora- 
tive cooling as well as by the requirements for gas exchange. Zeu- 
then (1942) and more recently Salt and Zeuthen( 1960) have suggested 
that the air sacs are important in producing the necessary evapora- 
tive cooling during flight. This suggestion appears plausible, but 
awaits experimental evaluation. 



The evolutionary history of those components of physical regulation 
affecting the extent of insulation inbirds is obscure, but it is possi- 
ble to delineate certain of the factors which must have influenced 
their development. Whether the evolution of feathers was originally 
related to heat conservation as suggested by Young (19 50) or to the 
establishment of capacities for gliding and ultimately, flight, it is 
obvious that both thermal considerations and aerodynamic require- 
ments have influenced their characteristics. The dual role of 
feathers appears to have imposed restrictions on the amount of vari- 
ation permissible in the thickness of plumage. Irving detected no 
major differences between arctic and tropical birds of comparable 
size in the thickness of plumage (Irving et al., 19 55). Subsequently 
he did note some minor structural differences between feathers of 
migratory and resident small birds in Alaska, which seemed to in- 
dicate that the latter had more effective insulation (Irving, 1960). 
In contrast to birds, large differences in pelage thickness between 
many arctic and tropical mammals are apparent (Scholander et al., 
19 50c). 

The aspect of physical regulation dependent on vasomotor ac- 
tivity could have been established in the avian line prior to the 
development of feathers as an outgrowth of the improvement of 
circulatory capacity necessitated by the intensification of metabol- 
ism. However, the character of the insulation afforded by the 
plumage probably provided a stimulus for the development of vaso- 
motor mechanisms to their present high level of performance in 
birds. Even with the inevitable wear and loss of feathers between 
molts, the minimum insulation afforded by the plumage is consider- 
able, and this must make heat loss from the naked or thinly feathered 
portions of the body of great importance during vigorous activity or 
hot weather. The thinly feathered sides are exposed by holding the 
wings away from the body in warm environments (Bartholomew and 
Dawson, 19 54b; Hutchinson, 1954). These areas may also serve as 
important sites of heat dissipation during flight. The unfeathered 
portions of the legs of various species appear to be important sites 
of heat dissipation under appropriate conditions (Bartholomew and 
Cade, 1957; Bartholomew and Howell, 1961). Combs and wattles of 
gallinaceous birds are apparently important in this respect also 
(see Yeates et al., 1941), although Hutchinson (19 54) does not agree 
that this has been convincingly demonstrated thus far. 



The beneficial role in the avian heat economy of vasomotor ad- 
justments favoring extensive blood flow through thinly feathered or 
naked regions of the body in warm environments is reversed in the 
cold and in most aquatic situations. This difficulty has been met by 
the development of means for restricting heat loss from them. Heat 
loss from the lower portions of the legs is apparently minimized in 
many species by curtailment of the blood supply (during inactivity 
they can also be protected by the body feathers when the bird "sits" 
on them). However, counter-current arrangements for heat exchange 
are evident in some species, for example, wading birds such as 
herons, cranes, and flamingoes (Hyrtl, 1863, 1864). In either case 
pronounced longitudinal temperature gradients can be produced. 
Irving and Krog (1955) found that leg temperatures in the Gull ( Larus 
glaucescens) ranged from 37.8 C proximallyto C distally when 
the animal was subjected to cold. Peripheral heterothermy, which is 
so important to the maintenance of centralhomeothermy, has appar- 
ently demanded thedevelopmentof mechanisms of temperature com- 
pensation m the peripheral tissues which are reminiscent of those 
occurring in poikilothermic animals (Bullock, 1955: and Fry, 1958). 
The demonstration of acclimation of conduction of the metatarsal 
portion of the peroneal nerve to cold in the Herring Gull ( Larus 
a rgentatus) provides an excellent example of this temperature com- 
pensation (Chatfield et al. , 19 53) . 

The temporal relation of the development of those components 
of physical regulation affecting the extent of insulation of birds to 
the actual appearance of homeothermy in the avian line is largely a 
matter of deduction. These components may have been present in 
incipient stages prior to the advent of this condition, but it seems 
reasonable that their full development occurred afterwards and was 
significant in conserving the increased amount of heat produced as 
a result of the metabolic changes discussed previously. If the fact 
that altricial birds (in which the events in the ontogeny of tempera- 
ture regulation can readily be observed because they occur after 
hatching) develop fairly effective control of body temperature through 
chemical regulation while their insulation is still in a rudimentary 
state (Pembrey, 1895; Ginglinger and Kayser, 1929; Baldwin and 
Kendeigh, 1932; and Dawson and Evans, 19 57, 1960) has any phylo- 
genetic significance, it would appear that the development of physical 



regulation has been mainly significant in reducing the energetic cost 
of homeothermy and in extending the range of environmental tem- 
peratures over which this condition can be maintained. Modification 
of this aspect of temperature regulation has, of course, subsequently 
comprised a major theme in the climatic adaptation of homeotherms. 

It is appropriate in connection with the evolution of physical 
thermoregulation to mention Bergmann's and Allen's Rules, which 
state that forms from higher latitudes tend to be larger and to have 
relatively smaller appendages than their counterparts from lower 
latitudes. The validity and significance of these rules in climatic ad- 
aptation have recently been the subjects of some controversy (Scho- 
lander, 1955, 19 56; Mayr, 1956; Newman, 19 56; and Irving, 1957). 
King and Earner's (1960:267) comments on these rules appear use- 

Neither of the rules appears to hold generally for most species 
with extensive latitudinal (and hence temperature) distributions. 
Furthermore, the relatively slight differences in the size of the body 
and length of the appendages are quite trivial with respect to adjust- 
ment of heat dissipation (Hutchinson, 1954; and Scholander, 1955, 
1956). This is not meant to argue for the invalidity of the "rules" in 
species in which such clines do clearly occur, for it is quite plausi- 
ble that these clines may have developed because of the slight en- 
ergy-conserving advantages conferred by these differences. It must 
be emphasized, however, that the magnitude of the changes in bodily 
dimension necessary to provide adequate ad justmentofheatdissipa- 
tion, or evenany appreciable fraction thereof, far exceeds the genet- 
ic potential of any species. 


Considerable variation in the state of development of birds at 
hatching is evident, and this is reflected indifference in thermore- 
gulatory capacities. At one extreme are the young of altricial spe- 
cies, e. g., passerines, which are hatched in a very immature state 
and do not develop effective temperature regulation until a week or 
more after emergence from the egg. At the other extreme are the 



young of precocial species, e. g., gallinaceous birds, which are 
hatched at a relatively mature state and soon afterward develop such 
regulation. Indeed, several observations indicate that some ability 
for temperature regulation is present in precocial birds even before 
hatching. Between the extremes represented by typically altricial 
and precocial birds are many species, e. g., cap rimulgids, which are 
intermediate in their developmental state at hatching. 

The precocial condition is assumed to be primitive in birds 
(Kendeigh, 1952). Evolution of the altricial condition has been keyed 
to the elaboration of patterns of parental behavior. Its appearance 
has been considered important from the standpoint of bioenergetics. 
The immature state ofnewly hatched altricial young and the relative- 
ly short period between fertilization and hatching allows a smaller 
egg of lower energy content than is generally found in precocial birds 
of similar adult size (Huxley, 1927). Therefore less demand is made 
on the energy resources of altricial females per egg produced. The 
fact that the young do not develop beyond a very immature state in 
the egg is compensated for by parental activity in their care and 
feeding. In many species this burden is shared by both parents. Par- 
ental behavior is effective in maintaining the young at near-homeo- 
thermic levels of body temperature before their powers of tempera- 
ture regulation become established. Thus they are able to develop 
under favorable conditions without having to expend energy beyond 
basic maintenance and developmental needs. In passerines, at least, 
the energetic obligations of homeothermy are only assumed when the 
young are nearing mature size (Kendeigh, 1939; and Dawson and 
Evans, 19 57, 19 60). On the other hand, precocial young, although they 
too may be brooded or may huddle with their siblings in cool envi- 
ronments (Lehmann, 1941; and Kleiber and Winchester, 1933) , must 
rely on their own energy to a considerable degree for growth and 
development and for maintenance of body temperature once they are 
hatched (Bartholomew and Dawson, 19 54a). 

It has been suggested (Kendeigh, 1952; and Witschi, 1956) that 
the evolution of small birds was in part made possible by the devel- 
ment of the altricial mode ofdevelopment. Certainly, the lower limit 



of avian size is reached onlyinaltricial species. The energetic con- 
siderations relating to smaller egg size and to the fact that the young 
are not required to assume the energetic obligauons of homeothermy 
during a major portion of their development support this suggestion. 


Any accountof the evolution of homeothermy in birds will neces- 
sarily be highly speculative on the basis of the information now at 
hand, but it can be useful in emphasizing what appear to be the prin- 
cipal determinants of this evolution and in suggesting pertinent lines 
of future research. The transition from poikilothermy to homeo- 
thermy in the avian evolutionary line must have involved may steps, 
some concurrent and some sequential. With the capacities for be- 
havioral control of body temperature which were probably present in 
the poikilothermic forms, this transition was probably more signifi- 
cant with respect to extension of the range of conditions over which 
body temperature could be held in the range suitable for activity than 
to any primary emancipation from the thermal environment. The 
initial steps in the establishment of physiological temperature regu- 
lation were probably metabolic, involving an overall intensification 
of metabolism with its far reaching demands on the structure and 
function of the various organ systems and then the development of 
chemical regulation with its underlying control mechanisms. Per- 
haps the rudiments of all the control mechanisms governing tem- 
perature regulation of birds were present in their poikilothermic 
antecedents , serving to control the behavioral and physiological com- 
ponents of the temperature regulation which these animals probably 
possessed. Extensive investigation of the neural mechanisms con- 
trolling thermoregulatory activity in birds and reptiles is needed be- 
fore an evaluation of this suggestion can be undertaken. 

The various components of the physical regulation in birds ap- 
parently became functional at different times. The presence of pant- 
ing in various contemporary thermophilic reptiles indicates that this 
process is not the sole property of homeo therms and raises the pos- 
sibility that it is present in birds as a legacy from their reptilian 



antecedents. It appears reasonable to postulate that the mechanisms 
controlling the extent of insulation were developed after consider- 
able quantities of heat became available to the nascent homeotherms 
with the intensification of metabolism. However, the bolstering of 
circulatory function required in this intensification must have pro- 
vided a preadaptation for physical regulation through establishment 
of an efficient heat transport system. The requirements of gliding 
and ultimately flight undoubtedly intervened in the development of 
that component of physical regulation involving the plumage. Conse- 
quently the role of the plumage as insulation is at least in part a 
compromise between thermal considerations and aerodynamic re- 
quirements. The nature of this insulation has undoubtedly lent great 
importance to the perfection of circulatory mechanisms having to do 
with control of heat loss from the thinly feathered and naked portions 
of the avian body. 

If the order of events in the establishment of birds was as sug- 
gested here, physiological temperature regulation must initially 
have been a costly process energetically. However, this would have 
been outweighed by the advantages which it conferred to the early 
homeotherm over its poikilothermic prey and competitors. Perhaps 
the diversification of the early homeotherms within the avian line 
contributed to a selection for the perfection of mechanisms of physi- 
cal thermoregulation; with this diversification competition among 
homeotherms would have been intensified, with the most efficient 
types having the advantage. 

Associated with the evolution of homeothermy in the avian line 
was the restriction of the thermal requirements for development. 
Satisfaction of these requirements was keyed to the development of 
patterns of parental behavior. It appears that theprecocial mode of 
development is the primitive condition inbirds. The evolution of the 
altricial mode of development has apparently been of great signifi- 
cance from the standpoint of bioenergetics because it requires a 
smaller egg of lower energy content than inprecocial development, 
because it restricts the utilizationof energy by the developing young 
to a minimum consistent with those maintenance processes exclusive 



of temperature regulation and with the requirements of development 
until the young are well on their way to mature size, and because it 
requires elaborate patterns of parental care following the hatching 
of the eggs. The evolution of small size in birds appears to have been 
contingent upon the development of the altricial condition. 

While the details of the evolution of temperature regulation in 
birds are obscure, they are less complex than those of mammals. 
Birds appear tobe a monophyletic group in which there was probably 
but one development of homeothermy. Mammals on the other hand 
are almost certainly polyphyletic (Olson, 1959; Simpson, 19 59), and 
homeothermy could have developed independently in each of several 
evolutionary lines as they traversed the boundary between mammal- 
like reptiles and mammals. Theprototherians (monotremes) and the 
therians (marsupials and placentals), the surviving groups of mam- 
mals, appear to have been separately derived from the mammal- like 
reptiles (Simpson, 1959), complicating considerations of the evolu- 
tion of homeothermy. 



1. Bailey, R. E. 1952. The incubation patch of passerine birds. Con- 

dor 54:121-136. 

2. Baldwin, S. P., and S. C. Kendeigh. 1932. Physiology of the tem- 

perature of birds Sci. Pub. Cleveland Mus. Hist. 3:1-196. 

3. Bartholomew, G. A., and T.J. Cade. 1956. Water consumption of 

house finches. Condor 58:406-412. 

4. Bartholomew, G. A., and T. J. Cade. 19 57. The body temperature 

of the American Kestrel, Falco sparverius. Wilson Bull. 69; 

5. Bartholomew, G. A., and W. R.Dawson. 19 52. Body temperature 

in nestling western gulls. Condor 54:58-60. 

6. Bartholomew, G. A., and W.R.Dawson. 1953. Respiratory water 

loss in some birds of southwestern United States. Physiol. 
Zool. 26:162-166. 

7. Bartholomew, G. A., and W.R.Dawson. 19 54a. Temperature re- 

gulation in young pelicans, herons, and gulls. Ecology 3 5:466- 

8. Bartholomew, G. A., and W.R.Dawson. 1954b. Body temperature 

and water requirements in the morning dove. Zenaidura mac - 
r oura marginella . Ecology 35:181-187. 

9. Bartholomew, G. A., and W. R. Dawson. 1958. Body temperature 

in California and Gambel's quail. Auk. 7 5:150-156. 

10. Bartholomew, G. A., W. R. Dawson, and E. J. O'Neill. 1953. A 

field study of temperature regulation in young white pelicans, 
Pelecanus erythrorhynchos. Ecology 34:554-560. 



11. Bartholomew, G. A., and T.R.Howell. 1961. Temperature regu- 

lation in Laysonandblack-footed albatrosses. Condor 63:185- 

12. Beddard, F. E. 1898. The structure and classification of birds. 

Longmans, Green and Co., London. 

13. Benedict, F. G. 1932. The physiology of large reptiles. ("Car- 

negie Institution of Washington Publications," No. 425) Car- 
negie Institution of Washington, Washington, D. C. 

14. Benedict, F. G. 1938. Vital energetics. ("Carnegie Institution 

of Washington Publications," No. 503) Carnegie Institution of 
Washington, Washington, D. C. 

15. Bogert, C. M. 1949a. Thermoregulation in reptiles, a factor in 

evolution. Evolution 3:195-211. 

16. Bogert, C. M. 1949b. Thermoregulation and eccritic body tem- 

peratures in Mexican lizards of the genus S celoporus . An. 
Inst. Biol. 20:415-426. 

17. Brooks, C. E. P. 1949. Climate through the ages. McGraw Hill 

Book Co., Inc., New York. 

18. Bullock, T. H. 19 55. Compensation for temperature in the meta- 

bolism and activity of poikilotherms. Biol. Rev. 30:311-342. 

19. Burton, A. C, and O. G. Edholm. 1955. Man in a cold environ- 

ment. Edward Arnold, Ltd., London. 

20. Chatfield, P. O., C. P. Lyman, and L.Irving. 1953. Physiologi- 

cal adaptation to cold ofperipheral nerve in the leg of the her- 
ring gull ( Larus argentatus) . Am. J. Physiol. 172:639-644. 

21. Cole, L. C. 1943. Experiments on toleration of high tempera- 

tures in lizards with reference to adaptive coloration. Eco- 
logy 24:94-108. 



22. Gowles, R. B., and C. M. Bogert. 1944. A preliminary study of 

the thermal requirements of desert reptiles. Bull. Am.Mus. 
Nat. Hist. 83:261-296. 

23 . Dawson, W. R. 19 54. Temperature regulation and water require- 

ments of the brown and Abert towhees , Pipilo fuscus and Pipi - 
lo aberti. Univ. Calif. Publ. Zool. 59:81-124. 

24. Dawson, W. R. 1958. Relation of oxygen consumption and evapo- 

rative water loss to temperature in the cardinal. Physiol. 
Zool. 31:37-48. 

25. Dawson, V/. R., and G. A. Bartholomew. 1958. Metabolic and 

cardiac responses to temperature in the lizard Dipsosaurus 
dorsalis. Physiol. Zool. 31:100-111. 

26. Dawson, W. R., and F. C. Evans. 1957. Relation of growth and 

development to temperature regulation in nestling field and 
chipping sparrows. Physiol. Zool. 30:315-327. 

27. Dawson, W. R., and F. C. Evans. I960. Relation of growth and 

development to temperature regulation in nestling vesper 
sparrows. Condor 62:329-340. 

28 . Dawson , W. R., and H . B . Tordoff . 19 59 . Relation of oxygen con- 

sumption to temperature in the evening grosbeak. Condor 61: 

29. Dill, D. B. 1938. Life, heat, and altitude. Harvard Univ. Press, 

Cambridge, Mass. 

30. Eisentraut, M. I960. Heat regulation in primitive mammals and 

in tropical species. Bull. Mus. Comp. Zool. 124:31-43. 

31. Eklund, C. R., and F. E. Charlton. 1958. Measuring the tem- 

peratures of incubating penguin eggs. O. N. R. Res. Rev., 
Dec, 1958. p. 1-6. 



32. Fitch, H. S. 1956. Temperature responses in free-living am- 

phibians and reptiles of northeastern Kansas. Univ. Kansas 
Publ. Mus. Nat. Hist. 8:417-476. 

33. Fry, F. E. J. 1958. Temperature compensation. Ann. Rev. 

Physiol. 20:207-224. 

34. Ginglinger, A., and C. Kayser. 1929. Etablissementdela ther- 

moregulation Chez les homeothermes au coursdudeveloppe- 
ment. Ann. physiol. et physicochem. biol. 5:710-758. 

35. Hart, J. S. 1958. Metabolic alterations during chronic exposure 

to cold. Fed.Proc. 17:1045-1054. 

36. Hemmingsen, A. M. 1960. Energy metabolism as related to body 

size and respiratory surfaces, and its evolution. Rep. Steno 
Memorial Hosp. Nord. Insulinlaboratorium 9(II):1-110. 

37. Hiestand, W. A., and W. C.Randall. 1942. Influence of proprio- 

ceptive vagal afferents on panting and accessory panting 
movements in mammals and birds. Am. J.Physiol. 138:12-15. 

38. Howell, T. R. 1959. A field study of temperature regulation in 

young least terns and commonnighthawks. Wilson Bull. 71:19- 

39. Huggins, R. A. 1941. Egg temperatures of wild birds under nat- 

ural conditions. Ecology 22:148-157. 

40. Hutchinson, J. C. D. 1954. Heat regulation in birds, p. 299-362. 

In J. Hammond, [edl. Progress in the physiology of farm ani- 
mals. (Vol. I). Butterworths Sci. Publ., London. 

41. Huxley, J. S. 1927. On the relation between egg- weight and body- 

weight in birds. J. Linn. Soc. 36:457-466. 



42. Hyrtl, J. 1863. Neue Wundernetze und Geflechtebei Vogein und 

Saugetheiren. Sitzber. Akad. Wiss. Wien. Abt. Ill, 48:6-7. 

43. Hyrtl, J. 1864. Neue Wundernetze und Geflechtebei Vogein und 

Saugetheiren. Denkschr. Akad. Wiss. Wien. 22:113-152. 

44. Irving, L. 1957. The usefulness of Scholander'sviewson adap- 

tive insulation of animals. Evolution 11:257-259. 

45. Irving, L. l960.Birdsof AnaktuvakPass, Kobuk, and Old Crow. 

U. S. Nat. Mus. Bull. 217, p. vii+409. 

46. Irving, L., H. Krog, and M. Monson. 1955. The metabolism of 

some Alaskan animals in winter and summer. Physiol. Zobl. 

47. Irving, L., and J. Krog. 1954. Body temperatures of arctic and 

subarctic birds and mammals. J. Appl. Physiol. 6:667-680. 

48. Irving, L., and J. Krog. 19 55. Skin temperature in the Arctic as 

a regulator of heat. J. Appl. Physiol. 7:355-364. 

49. Irving, L., and J. Krog. 1956. Temperature during the develop- 

ment of birds in arctic nests. Physiol. Zool. 29:195-205. 

50. Johansen, K. 1961. Temperature regulation in the nine-banded 

armadillo (D asypus novemcinctus mexicanus ). Physiol. Zo51. 

51. Johansen, K. 1962. The evolution of mammalian temperature re- 

gulation. This Symposium. 

52. Kayser, C. 1929a. Regulation thermique apres section medul- 

liare dorsole chez le pigeon. Gompt. soc. biol. 100:286-288. 

53. Kayser, C. 19 29b. Contribution a I'etudedume'canismenerveaux 

de la regulation thermique. Ann. physiol. physicochem. biol. 



54. Kendeigh, S.G. 1939. The relation of metabolism to the develop- 

ment of temperature regulation in birds. J. Exp. Zool. 82:419- 

55. Kendeigh, S. C. 1952. Parental care and its evolution in birds. 

Illinois Biol. Mon. 22:1-356. 

56. King, J. R., and D. S. Farner. 1960. Energy metabolism, ther- 

moregulation and body temperature, p. 215-288. In A. J. Mar- 
shall, [ed.]. Biology and comparative physiology of birds. (Vol. 
II). Academic Press, New York. 

57. Kleiber, M., and G. F. Winchester. 1933. Temperature regula- 

tion in baby chicks. Proc. Soc.Exper. Biol. Med. 31:158-159. 

58. Lehmann, V. W. 1941. Attwater's prairie chicken, its life his- 

tory and management. North American Fauna 57:1-63. 

59. Lohrl, H. 1955. Schlafengewohnheiten der Baumlalifer ( Certhia 

brachydactyla , C. familiaris) und andere Kleinvogelinkalten 
Winternachten. Volgelwarte 18:71-77. 

60. Martin, G. J. 1903. Thermal adjustment and respiratory ex- 

change in monotremes and marsupials— A study in the devel- 
opment of homeothermism. Phil. Trans. Roy. Soc. London, 
ser. B. 195:1-37. 

61. Mayr, E. 1956. Geographical character gradients and climatic 

adaptation. Evolution 10:105-108. 

62. Newman, M. T. 1956. Adaptation of man to cold climates. Evolu- 

tion 10:101-105. 

63. Norris, K. S. 19 53. The ecology of the desert iguana Dte^o- 

saurus dorsalis . Ecology 34:265-287. 



64. Odum, E. P. 1942. Muscle tremors and the development of tem- 

perature regulation in birds. Am. J. Physiol. 136:618-622, 

65. Olson, E. C. 1959. The evolution of mammalian characters. 

Evolution 13:344-353. 

66. Pearson, O. P. 1954. Habits of the lizard. Liolaemus multi - 

formis multiformis , at high altitudes in southern Peru. 
Copeia, p. 111-116. 

67. Pembrey, M. S. 1895. The effect of variations in external tem- 

perature upon the output of carbonic acid and the temperature 
of young animals. J. Physiol. 18:363-379. 

68. Rand ell, W. C. 1943. Factors influencing the temperature regu- 

lation of birds. Am. J. Physiol. 139:56-63. 

69. Rodbard, S. 1948. Body temperature, blood pressure, and hypo- 

thalamus. Science 108:413-415. 

70. Rogers, F. T. 1928. Studies on the brain stem. XI. The effects 

of artificial stimulation and of traumatism on the avian thala- 
mus. Am. J. Physiol. 86:639-650. 

71. Rogers, F. T., and R. W. Lackey. 1923. Studies on the brain 

stem VII. The respiratory exchange and heat production after 
destruction of the body temperatiire-regulating centers of the 
hjTJOthalamus. Am. J. Physiol. 66:453-460. 

72. Saafeld, E. von. 1936. Untersuchungen uber das Hacheln bei 

Tauben. Zeit. vergl. Physiol. 23:727-743. 

73. Saint-Girons, H., and M. C. Saint-Girons.TF56. Cycle d'acti- 

vite et thermoregulation chez les reptiles (lezards et ser- 
pents). Vie et milieu 7:133-226. 



74. Salt, G. W., and E. Zeuthen. 1960. The respiratory system, p. 

363-409. In A. J. Marshall, (ed.), Biology and comparative 
physiology of birds. (Vol. I). Academic Press, New York. 

75. Schmidt- Nielsen, K,, and B. Schmidt- Nielsen. 1952. Water me- 

tabolism of desert mammals. Physiol. Rev. 32:135-166. 

76. Scholander, P. F. 1955. Evolution of climatic adaptationin ho- 

meotherms. Evolution 9:15-26. 

77. Scholander, P. F. 1956. Climatic rules. Evolution 10:339-340. 

78. Scholander, P. F., R. Hock, V. Walters, and L. Irving. 1950a. 

Adaptation to cold in arctic and tropical mammals and birds 
in relation to body temperature, insulation, and basal meta- 
bolic rate. Biol. Bull. 99:259-271. 

79. Scholander, P. F., R. Hock, V. Walters, F. Johnson, and L. 

Irving. 1950b. Heat regulation in some artic and tropical 
mammals and birds. Biol. Bull. 99:237-258. 

80. Scholander, P. F., V. Walters, R. Hock, and L. Irving. 1950c. 

Body insulation of some arctic and tropical mammals and 
birds. Biol. Bull. 99:225-236. 

81. Simpson , G. G. 1959. Mesozoic mammals and the polyphyletic 

origin of mammals. Evolution 13:405-414. 

82. Stebbins, R. C, and R. M. Eakin. 1958. The role of the "third 

eye" in reptilian behavior. Am. Mus. Nov., no. 1870, p. 1-40. 

83. Steen, J. 1958. Climatic adaptation in some small northern 

birds. Ecology 39:625-629. 

84. Steen, J., and P. S. Enger. 1957. Muscular heat production in 

pigeons during exposure tocold. Am. J.Physiol. 190:157-158. 



85. Storer, R. W. 1960. Adaptive radiation inbirds.p. 15-55. In A. 

J. Marshall, ed. , Biology and comparative physiology of 
birds. (Vol. I). Academic Press, New York. 

86. Sullivan, C. M. 1954. Temperature reception and responses in 

fish. J. Fish. Res. Bd. Can. 11:153-170. 

87. Swinton, W. E. I960. Theoriginofbirds.p. 1-14. In A. J. Mar- 

shall, ed. , Biology and comparative physiology of birds. (Vol. 
I). Academic Press, New York. 

88. Templeton, J. R. 1960. Respirationand water loss at the higher 

temperatures in the desert iguana, Dipsosaurus dorsalis . 
Physiol. Zool. 33:136-145. 

89. Wallgren, H. 1954. Energy metabolism of two species of the 

genus Emberiza as correlated with distribution and migration. 
Acta zool. Fenn. 84:1-110. 

90. Wilhoft, D. C. 1958. The effect of temperature on thyroid his- 

tology and survival in the lizard, Sceloporus occidentalis . 
Copeia, p. 265-276. 

91. Witschi, E. 1956. Development of vertebrates. W. B. Saunders 

Co., Philadelphia. 

92. Yeates, N. T. M., D. H. K. Lee, and H. J. G. Hines. 1941. Re- 

actions of domestic fowls to hot atmospheres. P roc. Roy. Soc. 
Queensland 53:105-128. 

93. Young, J. Z. 1950. The life of vertebrates. Oxford Univ. Press, 


94. Zeuthen, E. 1953. Oxygen uptake as related to body size in or- 

ganisms. Quart. Rev. Biol. 28:1-12. 

95. Zeuthen, E. 1942. The ventilation of the respiratory tract in 

birds. Kgl. Danske Videnskab. Selskab Biol. Medd. 17:1-50. 



Kjell Johansen 

The term "evolution" has a special affinity to 
is an integrating symbol of what we are all coxicerned with. How- 
ever, most of us, I am sure, are somewhat reluctant to use the 
word in our scientific work. One reason for this, at least among 
physiologists, is that our section of biology is founded solidly on 
measurements and carefully designed experiments, and the infor- 
mation acquired does not readily fall into line with the observations 
that have formed the theories of evolution. Fossils, unfortunately, 
do not render themselves easily to physiological study. I have a 
strong interest In the possible routes along which physiological 
mechanisms may have evolved. In my near awe for the term "evolu- 
tion" I have found it expressive and also comforting to myself to say 
that I have an interest in the physiological phylogeny of certain func- 
tions. This term can be applied only to information compatible with 
the exactness required in a physiological study. Moreover, by using 
the phylum in a comparative manner, we are approaching the home 
grounds of evolution. As you can see from the program, Dr. Dawson 
and I have been ascribed the rather doubtful task of discussing the 
evolution of oneof the profound and striking physiological character- 
istics of thehighervertebrates,homeothermy.Today afew mamma- 
lian forms exist which retain a number of extinct morphological 
characters. These animals are often called living fossils or missing 
links, and are mainly represented by the Australian monotremes and 
marsupials. For many reasons, I have decided to confine the main 
parts of my discussion to these orders plus the New World marsupi- 
als and Xenarthra, which also represent the archiac forms rather 
well, in spite of their extreme specializations. Reference will also 
be made briefly to other orders of mammals classified among the 
more primitive forms. These are the Insectivores and the Chirop- 
tera, but there will be no time to discuss phylogenetic implications 
of the evolvement of temperature regulation within a distinct order, 
like for instance, the rodents. Brief digressions will be made to 
exemplify how and why environmental extremes may lead to evolve- 
ment of specialized physiological mechanisms also in the higher 



mammalian forms. The presentation will naturally have to be frag- 
mentary, and maybe more questions will be asked than answers 
given. This, I hope, will evoke a vigorous discussion in the distin- 
guished group of specialists present. Since I have selected the term 
"physiological phylogeny," I will make no or only superficial refer- 
ence to the important and intriguing problems related to the onto- 
genetic development of homeothermy in mammals. 

The paleontologist supplies us with some starting points that 
may be useful for our discussion. Our knowledge of the origin of 
mammals as it has been derived from fossils has been supplemented 
by certain surviving mammals which, in their morphology, indicate 
an early divergence from the main mammalian stocks. The mono- 
tremes can, with a fair degree of assurance, be traced back to the 
early Jurassic period, about 150 million years ago. On this basis, 
many writers have jumped to the conclusion that the mammalian line 
became warm blooded earlier than this date, probably as a response 
to changing climatic conditions, or by being driven by the dominant 
reptilian stocks to seek life in colder or warmer regions. 

The marsupials, showing striking similarities to the modern 
opossums, appear next in the fossil record. This indicates to us 
that the modem marsupials may be representative of the soft part 
conditions in mammals living 70 to 80 million years ago. A great 
many of today's marsupial features are, however, to be considered 
as specialized characters and not truly ancestral conditions. There 
are about 230 living species of marsupials, found mainly in the 
Austral- Asian regions, but there are also a few on the American 
continent. The marsupials show many similarities to placental mam- 
mals, particularly the Insectivores, which undoubtedly are the oldest 
stock of placental mammals. 

The placental mammals arose with the Insectivores in the 
Cretaceous period about 100 million years ago. All present living 
placental mammals have probably developed from these early 
Insectivores. The most archiac orders are the present living Insecti- 
vores, their close relatives, the bats (Ghiroptera), and a diverse 
group of Xenarthra. The bats, or Ghiroptera, show a very close 
morphological resemblance to the Insectivores, except of course, 
the specializations associated with flight. Among the Xenarthra 



(previously also termed Edentata) the armadillos (Dasypodidae) , 
have departed the least from the ancestral plan, and are a very 
ancient group, probably originating in the Paleocene, some 60 
million years ago. Both the anteaters and the sloths of today are 
decidedly very specialized animals. The remaining Xenarthra have 
left very scanty fossil evidence. To briefly complete our palfeon- 
to logical starting ground let us, via Figure 1, remind you of the 
remaining mammalian orders living today. Any paleontologist, and 
perhaps many of you, will be horrified at this unscrupulous sim- 
plicity and superficial treatment of mammalian descent. I feel, 
however, that this may be sufficient in our present context and 
prefer to return to more detailed descriptions only if they are advan- 
tageous to the physiological considerations to follow. 





cHiROPTERA — -Ill;;\\ A--::!!!!--*- UNGULATA 





Figure 1. Schematic arrangement of the existing mammalian orders. 

Turning then to the physiological phylogeny of temperature regu- 
lation, we should also list some starting points. First of all, it seems 
reasonable to assume that a certain variance in the environmental 
factors is necessary for the establishment of temperature regulation. 
In other words, before regulation appears there must be something 
to regulate against. This naturally leads us to believe that the grad- 
ual establishment of temperature regulatory mechanisms must have 



started under fairly constant environmental conditions. It seems 
also reasonable to assume that the first homeotherms were found 
among the terrestrial air breathers, because of the advantage of 
the low conductivity of the air to heat. It is true, of course, that 
a great many homeothermic animals, whales and seals, for example, 
are today found in the sea, but they are secondarily aquatic forms 
with particularly developed insulation. The most stable terrestrial 
conditions are found and have always been found in the tropics. This 
justifies the assumption that the first steps toward successful 
homeothermy were taken in the tropics. 

J. P. Darlington (1948) has advanced excellent arguments telling 
us that the animal dispersal both for poikilothermic and homeother- 
mic species started in the tropics and expanded north and south. 
This expansion and all migratory movements of animals are gen- 
erally very complex. Thus a successful species with a large dis- 
tribution range extending north and south can rarely be ascribed 
to one or a few characteristics. It seems, however, justifiable to 
assume that a rapid expansion and migration southwards and north- 
wards from the tropics must have had a bearing on a concomitant 
establishment of mechanisms for better regulation of internal tem- 
perature. You will see throughout this discussion, and, I am sure, 
in Dr. Morrison's paper as well, that the tropics today also have 
a great number of primitive forms. There we find the monotremes, 
most of the living marsupials, the overwhelming part of the Chirop- 
tera, the Insectivores, and practically all the Xenarthra except the 
nine-banded armadillo. 

Let us then start in the tropics and discuss the qualities known 
to us of the temperature regulating abilities of some of the primi- 
tive forms confined to this environment. We can realize that in 
order to make the transition from the reptilian to the primitive 
mammalian condition of temperature regulation, a rather radical 
change in the speed of the biological machinery had to take place, 
making possible a heat production high enough to keep a maintained 
high gradient of temperature between the core of the animal and 
the environment. The assessment of a higher internal temperature 
will in turn facilitate an accelerated nerve impulse, shortening of 
the latent period of a musclecontraction, and acceleration of diges- 
tion, among a number of other biochemical andbiophysically linked 



processes. All these features will result in an intensification 
of the life processes, making possible a greater exploitation of 
the environment. 

To keep a system at a constant temperature in a steady state 
situation, a change in heat loss must be balanced by a correspond- 
ing change in heat production, or vice versa. This requires a sen- 
sitive mechanism set to a particular temperature which exerts 
control over heat production or heat dissipation, or both. The sim- 
plest regulator we can think of in this respect would be what we 
are all familiar with in houses, refrigerators, etc., a regulator 
that turns on the heat if the temperature falls and shuts it off if 
the temperature rises above the set level. We can certainly appre- 
ciate the limitations of such a simple system. In order for it to 
work efficiently, the temperature of the object ought to be appre- 
ciably higher than that of the environment. Moreover, as the 
environmental temperature rises and exceeds that of the object, 
the whole system of regulation would fail. We shall soon see 
that this simplest possible system of regulation is exactly what 
we find in the lower mammals like the Echidna. We may ask why 
life did not choose the other possible way to achieve simple regula- 
tion, by regulating heat loss rather than heat production. An obvi- 
ous consequence of regulation of heat loss only would be a far 
higher fuel cost. Securing the necessary fuel for such a regulation 
would require both time and range of activity that were not pos- 
sible for the earliest mammals. The first records of body tem- 
peratures in monotremes were made about 75 years ago. In 1883 
Maclay published records of cloacal and intra-abdominal tempera- 
tures in two specimens of Echidn a aculeata. He found an average 
temperature of 28 C. Lendenfeld (1886) reported a marked increase 
(2 G) in the female Echidna after egg laying. Richard Semon (1894) 
seems to have made the first systematic study of body tempera- 
tures in monotremes on a fairly large number of specimens. The 
cloacal temperature he measured ranged between 26.5 C and 

34.0 C and intraperitoneally from 29.0 C to 36 .9 G. This repre- 

o o 

sents a fluctuation of 7 G to 7.5 G at air temperatures ranging 

only from 18 G to 24 G. Semon points to a clear intermediate pos- 
ition of the monotremes between the reptiles and the higher mam- 
mals, but he does not classify themaspoikilothermic, as was usual 



at that time. Semon expressed the hope that the monotremes would 
be just as important for the study of homeothermism in mammals 
as they had been to the study of comparative anatomy and develop- 
mental history. If Semon lived today, he would have been very much 
disappointed. In spite of a few very interesting studies that followed 
shortly after him, practically nothing has, to my knowledge, been 
done in the last 30 or 40 years , when the general study of tempera- 
ture regulation has flourished so greatly. Sutherland (1897) rep- 
orted 29.4 G to be the average temperature of 14 specimens of 
Echidna. One cold morning an animal could be as low as 22 C; 

whereas another one exposed to the mid-day heat registered as 

high as 36.6 G. This was to Sutherland an immense range for a 

mammal and suggested a reptilian lack of ability to regulate against 
temperature changes. Let me add that Sutherland did what practi- 
cally all of us do who study temperature regulation. He completely 
curtailed the animal's ability to regulate its body temperature by 
natural behavior. I hope to demonstrate repeatedly the importance 
of this factor. 

Maybe we should digress to put the importance of natural behav- 
ior in a proper relation to a phylogenetic discussion of temperature 
regulation. Let us then restate some of the essentials in the out- 
standing works of Gowles and Bogert (1944) on temperature regula- 
tion in terrestrial reptiles. The essence of their work is that terres- 
trial reptiles, that is, lizards and snakes, can and do keep remark- 
ably constant body temperatures during activity. Bogert (1949) 
introduces some clarifying terms when he refers to the birds and 
mammals as largely endothermic as opposed to the reptiles, which 
derive their body heat mostly from external sources and can thus 
be termed "ectothermic." The author points in particular to the 
importance of the solar radiation, which may raise a reptile's tem- 
perature to levels manydegrees higher than that of the air. It seems 
reasonable to accept the suggestion from Bogert that the acquisi- 
tion and perfection of the complicated machineryfor a high internal 
metabolic heat production has its antecedent in the ectothermic 
assessment of heat present today in the reptiles. The behavioral 
control of body temperature in reptiles implies a high degree of 
sensitivity to temperature changes. Sutherland's data seem to jus- 
tify the conclusion that the monotremes display variable body tem- 
peratures in response to a great range of air temperatures. In 1901 



to 1903, G. J. Martin did a very outstanding study on thermal adjust- 
ment in monotremes and marsupials. The sub-title of his paper was 
"A Study in the Developmentof Homeothermism."To my knowledge, 
this is the only published study with this title ever made. Since 
Martin, incontrastto his contemporaries, also tried to record meta- 
bolism and parameters of physical heat exchange and behavior, I 
prefer to postpone the discussion of his main findings until after a 
brief review of the following work done on temperature measure- 
ments of monotremes. In 1915, Wardlaw presented a long paper 
entitled "The Temperature of Echidna aculeata ." He reports average 
body temperature in the neighborhood of 30 G.Wardlaw's extensive 
records also contain data on seasonal as well as diurnal variations in 
body temperature. Morning temperatures would invariably be higher 
than afternoon temperatures, the difference being about 3 C. 
Wardlaw's data indicates a diurnal temperature change independent 
of the external air temperature. He also indirectly comments on the 
ability of the Echidna to increase rather rapidly its metabolism and 
body temperature during arousal from hibernation. 

Burrell, in his monograph on the platypus (1927), reports that 
the body temperature for seven females of this species ranged 
between 30° C and 33 C. Wood- Jones (1923) had earlier reported 
an average of 32.2 C for the same species. 

Martin (1903) set as his purpose locating the monotremes and 
marsupials on the ascending scale of superiority towards freedom 
from the environment. He recorded thebody temperature variations 
at controlled air temperatures between 4 C and 40 C. His main 
concern was, however, to ascertain to what extent variation in heat 
production and variation in heat loss were used for purposes of 
adjustment. For comparison, he used lizards, Cyclodusgigas,cats, 
and rabbits besides the monotremes and marsupials. Figure 2 rep- 
resents Martin's findings in regard to the relationships between 
body temperature and air temperature. The results were recorded 
under laboratory conditions where behavior as a means of adjust- 
ment was drastically reduced or impaired. The Echidnas display a 
variance of about 10 C between the extremes of air temperature. 
Ornithorynchus displays somewhat less variance and also regulates 
at a higher level. The marsupials studied, Dasypus maculatus , 
Bettongia, (the kangaroo rat) and the opposum, Trichosurus 








-— — "• 

Cat • ' 

Rabbtt •- ' 



"^ ■■■■'"' 


nrnilhnrhvnrhiK ^ 


Ectinida ^■^ 




Cyclodus / 

10° 20° 30' 



Figure 2. Relationship between body temperature and ambient temperature 
for the lizard (Cyclodus sp.) and the monotremes, marsupials, and placental 
mammals studied by C. J.l^artin (1903). The values for Bradypus from Sawaya 



o o 

v ulpecula , showed a variance of about 3 C--between 36.1 C and 

SS.e*^ C. His cats showed a variance of 1.4 C whereas the rabbits 
showed a range of 3.6° C. Martin did what very few people do but 
what is necessary to get intelligible results; he reported the expo- 
sure time to the various air temperatures, and also regulated air 
temperature both up and down while recording body temperature. 
His exposure times ranged between 60 and 100 minutes. Martin 
argues that the platypus has beenunjustly listed as a poikilothermic 
animal. Between 5 G and 30 C air temperature, he claims that it 
adjusts its body temperature even better than the rabbit. However, 
when subjected to a temperature above 30 G, it became what Martin 
called "feverish." Observations done by a number of naturalists on 
the platypus indicate that its relation to the aquatic medium may be 
important for its temperature regulation. By a closer examination, 
it becomes apparent that body cooling is frequently attained in the 
tropics by returningto water (Hesse, 19 37). This is particularly true 
among larger animals, like the water buffalo, the water buck, the 
rhinoceros, the elephant, and most strikingly, the hippopotamus. The 
elephant also frequently operates his personal, built- in shower, and 
his ears are of paramount importance in temperature regulation. 
These animals have a number of interesting specialities in their 
temperature regulation. However, the discussion of these I think be- 
longs more properly in Dr. Morrison's paper. For the purpose of 
my discussion, it suffices to emphasize that body cooling by return- 
ing to water frequently occurs in tropical forms and also among the 
more primitive ones. Gooling results not only from staying sub- 
merged and benefiting from the larger conductivity of the water for 
heat, but also from frequent emergence from the water and obtaining 
a cooling effect by way of the evaporative characteristics of the air. 
Significantly, a great percentage of the primitive forms aid their 
less-developed temperature regulation by burrowing. Itis interest- 
ing that the substratum temperatures in tropical Australia never 
exceed 85 F (Vorhies, 1945). 

Martin's data on metabolic heat production shows a number of 
interesting features. Minimum heat production was found at about 
30 G in all the species, including the higher mammals. His fig- 
ures, calculated according to body surface show a similar metabolic 
rate for both the monotremes and the marsupials, which were only 
one-third of the values he got for the higher placental mammals. 


^ .8H 

Bettongio i 

?N ■■•-.. V ••"" 

10° 20° 30' 



Figure 3. The metabolic values recorded by C. J. Martin (1903) in his studies 
on monotremes and marsupial and placental mammals. 



This low level of metabolism in these primitive forms is, of course, 
of the greatest interest. Martin found about 2.5 times increase in 
metabolism of Echidna at environmental temperatures of 5 C. He 
points out that none of the experimental animals he used maintained 
a constant body temperature throughout the range of air tempera- 

Of great significance to our understanding about how homeo- 
thermy has developed phylogenetically is the finding that monotremes 
depend only on variation in heat production and not on physical mech- 
anisms for their maintenance of body temperature. At high air tem- 
peratures, that is about 30° C,the respiratory rates of the marsup- 
ials are affected very little by the high temperatures; whereas in 
the monotremes, the breathing frequencies are decidedly lower in 
30° G air than in cooler air. Martin was unable to find any sweat 
glands in Echidna , and he also demonstrated that this species is de- 
void of vasomotor adjustments importantforheatdissipation or con- 
servation. Without any means for adjusting the core temperature by 
physical means, Echidna is vulnerable at high air temperatures and 
dies easily of heat apoplexia at a body temperature as low as 38 C. 
Under natural circumstances, I would assume that behavior mecha- 
nisms are indispensible to Echidna. The animal is known to bury it- 
self several feet in the ground and only emerges after sundown on hot 
days. This, in turn, willdrastically curtail the animal's activity time 
and range. 

Kathleen Robinson, in 1954, studied heat tolerance in Australian 
monotremes and marsupials. She confirmed Martin's earlier find- 
ings about the lackofpanting and vasomotor adjustment in the mono- 
tremes. She measured the evaporative heat loss and found that it is 
higher in platypus than Echidna, indicating some activity of the sweat 
glands in the former. These are distributed mainly on the snout, but 
there is also one apocrine gland opening into the follicle of each of 
the large hairs. The platypus shows also some adjustment in posture 
to facilitate heat loss. As its body temperature rises, it rolls over to 
its back with the under surface exposed and the legs outstretched. 
With higher body temperatures, the platypus becomes restless, and 
indicates some impairment in neuro- muscular coordination. The 
animals do not salivate or lick their coats during heat exposure. On 
the basis of the scattered data presented so far, we may be justified 


in making a first attempt to characterize temperature regulation in 
monotremes. These animals, the most primitive among mammals 
available to us for experimentation, are definitely homeothermic; 
that is, they can regulate against environmental changes at a level 
higher than that of the air . This is achieved by variance in heat prod- 
uction and behavior. At air temperatures around and above the body 
temperature such regulation fails, and homeostasis is maintained by 
behavior. This primitive condition requires a central nervous inte- 
grating control of both behavior and heat production. Understandably 
these animals are confined to tropical stability and seek refuge in the 
stability of the substratum at the high air temperatures. The other 
member of the living monotremes, Omithoryhchus (platypus) , shows 
some advance in regulating ability over Echidna by having sweat 
glands . 

The marsupials show a distinct advance in homeothermism, be- 
ing able to vary both heat production and heat loss. I will briefly 
refer to some of the work done on temperature regulation in mar- 

Sutherland (1897) observed body temperatures on sixteen dif- 
ferent species of marsupials. The average body temperature for all 
the species was 36 G. Sutherland lists the wombat ( Phascolomys 
plathyrinus) as the poorest regulator with an average temperature 
of 34 C. Next came the members of the genus Petaurus or the fly- 
ing squirrels, with an average of 35. 7 C, The koala bears (Phasco- 

larctos cinereus) had a range of from 35.0^ C to 36.5° C at air tem- 

o o 

peratures between 7.7 C and 24.5 C. Sutherland adds that upon 

exposure to the sun the body temperature rose rapidly, in one speci- 
men to 38.4 C. Very interestingly, females always showed higher 
temperatures than males under the same conditions, and the diver- 
gence was always greatest when the females were suckling their 
young. The average excess temperature from 25 observations was 
1.2 C. The fact that pregnant or suckling females have higher tem- 
peratures and also a more efficient regulation has often been report- 
ed, for example, Morrison (1945) for a pregnant sloth. In the larger 
marsupials, notably the kangaroo ( Mac r opus giganteus ) tempera- 
tures between 36 G and 37 G were recorded. Some marsupials. 



according to Krieg(1952), hibernate or have a similar torpid con- 
dition. In Marmosa cinerea . which is considered old from a phylo- 
genetic standpoint, Eisentraut (1955) recorded an average body tem- 
perature of 34.7° G, and a range of from 29.3 C to 37.8 G, show- 
ing great dependence upon conditions of activity. I would like to 
give Eisentraut credit for his attempt at systematically arranging 
body temperatures, not as fixed numbers, but as ranges of tem- 
peratures. To indicate an animal's body temperature as a fixed 
point, even when giving this as an average, is not as expressive 
as listing the range in body temperature for the animal during 
normal natural conditions. Such information requires undisturbed 
recordings of body temperatures under all normal conditions rang- 
ing from sleep to strenuous exercise. Information of this kind is, 
unfortunately, available for only a very limited number of species. 
Such information, however, will, in my mind, express more about 
an animal's temperature regulating ability than do most of the para- 
meters now in general use. Eisentraut (1956b) suggests "activity 
temperature" as a term for such a range. This apparently excludes 
conditions of sleep and rest, which ought to be included. I will, 
therefore, propose to call it "body temperature range at normal 
behavior." It is highly significant that the New World Didelphidae 
(Didelphis, Marmosa) have lower body temperatures than the mar- 
supials of the Old World. The taxonomists unanimously consider 
the New World marsupials as the most primitive phylogenetically. 
Morrison, in his work on two marsupials from Central America 
(1956), states that the species studied, the brown opossum ( Meta - 
chirus nudicaudatus ) and the Eten opossum ( Didelphis marsupialis ) , 
showed a homeothermism in no way inferior to that of many higher 
mammals. Sutherland's data (1897) show that even the very slug- 
gish koala bear (Phascolarctos) maintained his body temperature 
better than any of the placental mammals. A number of authors 
comment on the well-developed ability of the common American 
opossum ( Didelphis virginiana) to regulate its body temperature. 
As we all know, this species has migrated extensively northward, 
and it seems of the greatest significance that its temperature regu- 
lation is decidedly superior to that seen in the closely related 
tropical forms. 



Robinson (1954) studied the development of the mechanisms 
for evaporative heat loss in Australian marsupials. She found a 
close correlation with structural evolutionary trends. She recorded 
breathing rate, pulse rate, evaporative weight loss, and she also 
studied sweat patterns. There is a great amount of evaporation 
from the respiratory tract and additional evaporation from the 
buccal mucosa during open-mouthed panting. Sweat glands were 
easily located over the entire body surface (Bolliger and Hardy, 
1945). The sweat glands are, however, of a primitive apocrine type 
and of seemingly little importance as evaporative mechanisms. 
Evaporation is, on the other hand, significantly aided by salivation 
and coat licking. Robinson concludes that the heat tolerance in the 
Australian marsupials studied followed the ascending order of phylo- 
genetic development, for instance, the primitive bandicoot (Peram- 
eles n acuta ) , next the opossum (Trichosurus caninus ), then the 
cuscus ( Spilocuscus nudicaudatus) , the koala ( Phascolarctos cine- 
reus), the wallaby ( Petrogale penicillata ) , and most superior, the 
wallaroo ( Mac r opus robustus). 

Higginbotham and Koan, in 1955, studied temperature regula- 
tion at elevated air temperatures in the Virginia opossum ( Didelphis 
virginiana) . They found that when body temperature increased to 
about 38 C, panting, profuse salivation, and licking of saliva upon 
feet and tail and parts of the trunk, was common. They monitored 
anaesthesia to a point where panting and salivation still persisted, 
but coat licking, of course, was abolished. This fact prevented the 
animal from keeping the body temperature at sub- lethal levels, 
and they conclude thatthe spreading of saliva upon the body surfaces 
and subsequent evaporation constitutes an indispensable mechanism 
for heat dissipation. Robinson and Morrison (19 57) studied the reac- 
tions to hot atmospheres of various species of Australian marsu- 
pials and placental mammals. Their material covers as many as 
25 species of Australian marsupials, plus 4 indigenous Australian 
rodents. They make the very interesting and, in my mind, important 
comparisons of temperature response to activity in some of their 
subjects. In members of the Dasyuridae, rises of 4 C were not 
uncommon. They report that maintenance of body temperature at a 

constant but higher than normal value was successfully achieved by 

° o o 

all their animals at air temperatures of 35 C. At 40 C air tem- 
peratures, some species failed to adjust to a steady state condition. 



However, all theirspeciesofPhalangeridae achieved thermal equili- 
brium at an air temperature of 40 C. The carnivorous marsupials 
seemed to be particularly vulnerable at the 40 C air temperature. 
Most of the species showed increased respiratory rates at the high 
air temperatures. Also open- mouth panting, and salivation and coat 
licking in typical postures promoting heat dissipation frequently 
appeared. The Tasmanian devil ( Sarcophilus harisii), was an 
exception and showed no reaction other than an increase in water 
consumption. The authors point to the fact that the species unable 
to maintain equilibrium at 40 C were all among the most phylo- 
genetically primitive. The author's attempt to classify the ability 
among mammals to dissipate heat is shown in Tables I and II, 
which are taken from their work. 

Bartholomew (1956) has presented perhaps the only detailed 
study of the various facets of temperature regulation in a marsu- 
pial. He made careful studies of the macropod(Setonix brachyurus ) 
both under laboratory conditions and in the field. He recorded a 
considerable diurnal lability in body temperature related to the 
daily cycle of activity in the field. The species studied showed a 
typical nocturnal activity pattern, and the day-time rectal tempera- 
tures of 37 C were significantly lower than the night-time tem- 
peratures. It is important that the slight excitement occurring dur- 
ing attachment of thermocouples could cause temporary elevations 
of body temperature up to 1.5 C in the rectal temperature. Upon 
the exposure to high air temperatures (40 C), a copious secretion 
of saliva and licking of the feet and tail, and a distinct vasodila- 
tion of peripheral parts seem responsible for maintenance of ther- 
mal balance at these high air temperatures. The increase in res- 
piratory rate was appreciable, up to 200 per minute, but never as 
vigorous as, for instance, panting in a dog. When the animals were 
returned to room temperatures of 20 C there were indications 
that the peripheral vasodilation persisted for some time after 
removal of the heat stress. The elevation of the deep-body tem- 
perature during the heat load seemed moderate and did not exceed 
1 G. At extreme heat stress with dry bulb temperatures of 44 C 
for 4 hours or more, the animals showed no apparent failure to 
maintain the thermal equilibrium (Fig. 4). Temperatures of periph- 
eral parts like the feet and tail, rose rapidly to levels almost 






Phalange ridae 



Garni vora 
Prima ta 



Mars. Mice 
Mars. Rat 


Thermal Equilibrium 
AT > 2° C 2° C >Ar > 1° C AT < 1° C 

Giant naked 
-tail rat 
Water rat 


Nat. Cat 




White mouse 

White rat 

Fruit bat 




Rat. Kanga. 




Tas. Devil 


Field rat 




Table I. Comparison of the heat-regulatory ability of mammals at 40° C. 
From Robinson and Morrison, 1957. 



Normal Rate 

Modification at 40 C 
Slight Marked 

Slow Echidna Bandicoot 

(20-30/min) Platypus 


Moderate Tas. Devil Gliders 

(60-120/min) Cuscus Possums 

Water rat Koala 

White rat Wallabies 

Monkey Wallaroo 

Fruit bat 


Fast Sugar glider Mars, nnice 

( ^150/min) White mouse Mars. 

Naked-tail rat Nat. cd 

Field rat 
Mars, rat 

Table II. Comparison of respiratory response in mammals at 40° 
C. From Robinson and Morrison, 1957. 






J I 1 L 


20 40 60 80 100 120 140 160 180 200 220 240 

Figure 4. The effects of a dry bulb air temperature, of 44 C on the body tem- 
perature of an adult Setonix sp. (G. A. Bartholomew, 1956). 



within the range of the deep body temperature. The foot and tail 
were, however, held at levels below the core and the environment 
by conspicuous secretion and licking with saliva. Again, upon a 
sudden change to 20 G air temperatures, the maximally involved 
heat loss mechanisms showed some delay in adjusting to the new 
conditions, and body temperatures fell rapidly 2 C to 2 1/2 C. 
The licking and salivation seemed, however, to stop immediately 
upon cessation of the heat stress. At moderately low air tempera- 
tures (3 C) two animals averaged slightly lower body temperatures 
than normal for quiet animals at room temperature, whereas, one 
animal over-compensated and showed a body temperature above 
resting levels for most of the time. All animals shivered violently 
during the cold exposure. Again I think it is highly important that 
the body temperatures of the animals rose rapidly after removal 
from the cold (Fig. 5), indicating again some persistence of the 
compensatory mechanisms, this time violent shivering after the 
actual cessation of the cold stress. Bartholomew tried to evaluate 
the peripheral vascular situation by temperature measurements. 
His data, as in Figure 6, shows very conspicuous gradients, in 
particular along the tail. Bartholomew interprets the gradients 
in proportion to the degree of local vasoconstriction. It seems, 
however, conceivable that a counter- current effect, which in other 
animals is responsible for tremendous temperature gradients in 
peripheral extremities, might be in operation and thus conserve 
heat. Again, Bartholomew noted that there is a gradual diminution 
in peripheral vasoconstriction following exposure to cold. In my 
own interpretations of his data these repeated findings of delays 
in the transitions between regulatory states indicate some slow- 
ness in the integrative apparatus controlling the effector responses 
in temperature regulation. I will try to elaborate this assumption 
in more detail when discussing my own data on the armadillo. 
Bartholomew also exposed some wallabies to -10 Gfor two hours. 
Such temperatures are never encountered by this species in its 
normal habitat. At this time, both the subjects studied over-compen- 
sated to the cold stress, and showed a deep body temperature more 
than 1 C above the normal resting level. The feet and tail showed 
a marked vasoconstriction, and the peripheral temperature ap- 
proached C. Of great interest, the foot, and to a lesser degree 
the tail, showed waves of vasodilation very similar to cold vaso- 
dilation. Bartholomew's data indicate clearly that the ability to 




Figure 5. The responseofdeepbody temperatures of three specimens of Setonix 
sp. subjected to prolonged exposure to air temperatures as low as they ever en- 
counter under natural conditions. (G. A. Bartholomew, 1956). 





<-> 25 


/ > ..^- 


2.0-3.2° — )+t 


, , I , . , I , , . I . . . I . . . I ■ ■ ■ I ■ ■ ' I ■ ■ ' I ■ ■ ■ I ' ' ' ' ' ' ' I ' ' ■ 
20 40 60 80 100 140 180 220 


Figure 6. The deep body and skin temperatures of an adult Setonix sp. during 
exposure to low air temperatures. (G. A. Bartholomew, 1956). 



regulate deep body temperature in the marsupial (Setonix sp.) is 
equally as efficient as that displayed by many placental mammals 
of similar size. The ability to regulate even extends to air tem- 
peratures below and above those ever encountered in the animal's 
natural habitat. Under' severe positive heat loads, both sweating 
and panting was decidedly less important than copious salivation 
and licking. This seems to be a general specialization within the 
marsupial order; it has been demonstrated by Robinson and 
Morrison (1957), and others. Bartholomew points out that this 
very effective mechanism for facilitating heat loss is a specialized 










'■4« AIR 

20 40 60 80 100 120 140 160 180 

Figure 7. The response in body temperature o£ an adult Setonix sp. to an air 
temperature much low^r than that which ever normally occurs m its environment. 



behavioral response, in contrast to the pure physiological mechan- 
isms of sweating, panting, and vasodilation. This specialization 
limits the usefulness of the method since it can only operate effect- 
ively in an animal resting, and would be less useful to a rapidly 
moving animal. To briefly summarize the temperature regulating 
ability of the marsupials living today is very difficult, if not impos- 
sible. The group is exceedingly diversified, and having been prac- 
tically indigenous to Australia with little competition for a very 
long time, marsupial Life has adjusted to most habitats available. 
With the very interesting exception of the Virginia opossum of 
North America, the marsupials are confined to the tropical or 
neotropical regions. This, however, may not be related so much 
to inferiority to placental mammals in temperature regulation 
as to their very specialized mode of reproduction. In this regard, 
Bartholomew presents very interesting and important data on the 
ontogenetic development of temperature regulation in the marsu- 
pial, Setonix sp. I consider this a topic in itself, however, and can 
find no time to discuss it now. The marsupials so far studied in 
regard to temperature regulation indicate clearly a lower level 
of resting body temperature in the more primitive forms like the 
Dasyuridae which show values down to 33 G to 34 C at resting 
conditions, at 20 G air temperature; whereas the specialized, 
phylogenetically more advanced species show a higher resting 
level and a smaller range of variation. As more information be- 
comes available, like the important works on Australian marsu- 
pials by Robinson and Morrison (1957) and Bartholomew (1956) , this 
unique indigenous fauna may enable us to talk with more confidence 
about the role of the phylogenetic position vis a vis the influence 
of environmental factors for the establishment of homeothermy. 
Common to temperature regulation in all marsupials is the presence 
of physiological effector mechanisms of both chemical and physi- 
cal temperature regulation. Some of them, like sweating and pant- 
ing, seem generally to be of rather limited importance, being sub- 
stituted by specialized behavioral responses like the coat licking. 
In my own interpretation of Bartholomew's important work on 
Setonix, it seems of the greatest significance that the integrative 
control of the otherwise well-developed effector mechanisms 
show some lag in precision, compared to higher placental mam- 
mals. It is conceivable that even more striking differences in this 



respect are demonstrable in the more primitive marsupials not 
yet subjected to such a detailed examination. 

Turning next to the placental mammals, we are, of course, 
confronted with an even greater complexity in phylogenetic develop- 
ment and diversity in ecology than for the marsupials. The limita- 
tion deemed necessary in this treatment may reflect a personal 
bias, and I do hope the subsequent discussion will give room for 
your feelings about these problems. 

A starting point for the phylogenetic discussion of temperature 
regulation among the placental mammals has to be the Insectivores. 
The tenrec of Madagascar ( Gentetes ecaudatus) is probably the 
most phylogenetic any primitive of all the placental mammals liv- 
ing today. Eisentraut (1955, 1956b), who together with Rand (1925), 
seems to be the only worker having experimented with temperature 
regulation in this important species, states that the animal estivates 
during the dry season, which corresponds to the winter season. The 
tenrec shows in general an extraordinary labile body temperature, 
fluctuating between 24.1 C and 34.8 C. The tenrec is a typical 
nocturnal animal and shows a diurnal cycle of more than 10 C 
when the air temperature changes only 3 C to 4 G. The body tem- 
perature must thus be closely related to the activity of the animal. 
Eisentraut notes that the animal is able to perform normal coordin- 
ated activity at body temperatures down to 25 G. It is unfortunate 
that no detailed study measuring other parameters than body tem- 
peratures has yet been done on this very interesting species. Rand 
comments briefly that two other species of tenrec s ( Hemicentetes 
semispinosus and Setifer setosus) remain active all year around. 

Eisentraut (1956b) also reports body temperatures in two other 
species of primitive Insectivores ( Hemiechinus auratus and Parae - 
nhiniiR aethiopicus) . They show a range in body temperature of 
33.4° G to 36.4° G and 31.2 C to 36.2 G respectively. We may 
briefly comment on the extensive studies of the European hedge- 
hog, particularly referring to its hibernating ability. Most authors 
report a rather large lability in the body temperature of this species, 
from 31.2 G to 36.5 G. Morrison (19 57) has argued that this great 
range in all probability results from measurements taken in the 



hibernating season when the animal is in a transitory state. When 
only body temperatures from representative times of activity are 
taken, Herter (1934) reports a range of 2 C, and reports a dif- 
ference in activity temperature between summer and winter of 
about 1 G for the hedgehog. The one specimen of Erinaceus 
europeus subjected to measurements by Morrison (1957), shows 
the range of 34.8 C to 36.4 C with an average of 35.6 C. Although 
it has been decided that we leave out any detailed discussion of 
hibernation in this symposium, it seems justifiable to comment on 
Eisentraut's statement (1956a) that hibernators generally show 
imperfect heat regulation also during the active season; and he 
includes all known hibernators in the group of lower warm-blooded 
animals. I am, myself, and I know many others are, willing to 
challenge this statement. "Hibernation," although used to express 
the seasonal and diurnal condition of sleep displayed by a variety 
of the smaller mammals is really not representative as a term 
for all these species, and a great number of the true hibernators 
can, in my mind, be classified as extremely specialized and very 
far from primitive in their mode of temperature regulation. 

The shrews represent an interesting and successful group of 
Insectivores. In spite of their small size, which is obviously dis- 
advantageous in the cold, they have left behind most fellow Insecti- 
vores and invaded the north temperate and even arctic regions. 
They are, for instance, found here aroiind Fairbanks and in Norway 
even farther north. Apparently, no detailed studies have been made 
of these extreme northern populations of shrews. Morrison, Ryser, 
and Dawe (1959) have, however, presented a careful study of the 
shrew (Sorex cinereus) obtained from the Wisconsin region. The 
authors noted that the manipulation and handling of the animals dur- 
ing measurements of their body temperatures invariably increased 
the body temperature. The increase occurs very rapidly, often 
1 G per minute, and the body temperature could attain a level 
above 41 G. Similar elevation in body temperature became apparent 
after exercise. The body temperature seems always to level off 
at about 41 G. Figure 8 shows the results. More than anything it 
demonstrates a great variability of body temperature; this animal 
is unquestionably one of the more successful small animals living 
today. Its labile body temperature can hardly be classified as a 



Nl 3dniVd3dl/\l31 AQOa 

'-' a> M (-I 

S C CO ^ 

2 c (h o 

■^ ^ s 

5 M o <" 



















































tH g 0) 

C --^ m bo 
■2 to -£ C 





















• - 
















































<H > " 


° I' ■;;? 


<u " p 


s ■§ § 





s -» t5 


5 oT^ 

o o 

>> 3 

ffl 2 

C O 

• . rt E 

■« a^ ■S? P 


c > 

CO)* • 
cd 3 ^ bo 



primitive characteristic. It seems obvious to me that an animal 
this size together with a number of other smaller ones, like for 
instance the birch mouse (weighing 7 to 12 grams) I have studied 
(Johansen and Krog, 19 59), and the little pocket mouse ( Perognathus 
longimembris ) studied by Bartholomew and Cade (1957), have to 
allow for a greater variability in the body temperatures to regulate 
at all. We can appreciate the enormous activity needed by these 
animals to secure enough fuel for their high- paced metabolic 
machinery. The labile body temperature is thus in a number of 
species expressive of a specialization rather than a primitive 
character. It should prove most interesting to study temperature 
regulation in shrews of the northern-most habitats. 

The exceedingly specialized bats, the Chiroptera , offer a num- 
ber of interesting features in mammalian phylogeny. They are 
mostly confined to the tropics, although some small Micro- 
Ghiroptera, Vespertilionidae, approach the arctic regions on the 
Scandinavian Peninsula and in Alaska. It seems to the benefit of 
all of us, however, that I leave the discussion of this important 
order to a far greater specialist than myself. Dr. Morrison, who 
has done extensive research on a number of species both among 
the Micro- and Mega- Chiroptera (Morrison, 1959). Before I 
leave the important Insectivores, let me remind you that they are 
generally small in size, mostly confined to the tropical and tem- 
perate regions. Theydisplay a rather great lability in body tempera- 
ture, and many of them are hibemators. A substantial number of 
species are typical substratum dwellers, with burrowing habits. 
Some are excellent nest builders. Their body temperatures tend to 
show a large range and lability, but are regulated at a higher level 
than those of most marsupials. 

Let us next turn to the phylogenetically very interesting group, 
Xenarthra. They consist of the armadillos, the Dasypodidae, the 
anteaters, the Myrmecophagidae, the sloths, the Bradypodidae, and 
the Old World scaly anteaters, the Manidae. The extremely inter- 
esting and phylogenetically important Orycteropus called the 
aardvark, or Cape Anteater, is a zoologically very isolated form. 
This animal retains a number of characteristics present in the 
earliest eutherians. Some authors, like H. Winge (1941) place it 



together with the edentates (Xenarthra) , others with the insecti- 
vores, and still others at the base of the ungulate stock. Unfor- 
tunately for us, nothing has been done yet with this unique repre- 
sentative of the primitive placental mammals. 

Representatives of the Old World's scaly anteaters have like- 
wise scarcely been subjected to study in regard to their tempera- 
ture regulating ability. Eisentraut (1956b^ reports a body tempera- 
ture range for Manis tricuspis from 32.2 C to 35.2 C. A similarly 
limited number of observations is available on the tropical anteaters 
from Central America. WislocM and Enders (1935) report that the 
giant anteater ( Myrmecophaga jubata ) shows a rectal temperature 
between 32° C and 34 C at air temperatures between 16 C and 
21° C. The three- toed anteater ( T am andua tetradactyla ) shows rec- 
tal temperatures between 33.7 C and 34.6 ^C at air temperatures 
between 25° C and 27.6° C. The more sluggish two- toed anteater 
(Cyclopes didactylas) displayed a lower level of body temperature 
between 28^.9'' C and 31.3 C and showed a greater variance during 
exercise than the other two species. Enders and Davis (1936) 
recorded somewhat higher rectal temperatures on the T am andua 
tetradactylum . They got 35.0° C to 35.7 C at an air temperature 
of 27 C. The uniquely specialized group of sloths has been studied 
by a number af workers. In 1924 Ozorio de Almeida and Branca 
de A. Fialho observed a range from 30 C to 32.9 C at an air 
temperature between 19 C and 25.8 C for the three-toed sloth 
( Bradypus griseus ). Kredel (1928) recorded a range from 27.7 C 
to 36.8 C at air temperatures from 24.5 C to 32.4 C. Upon 
exposure to a moderate cold stress, the three-toed sloth loses 
considerable heat rapidly. According to Gibbs, cited by Wislocki 
(1933), Bradypus griseus lost 8 C in deep rectal temperature, 
dropping from 33 C to 25 C in 2 hours and 40 minutes when 
transferred from an air temperature of 26 C to 13 C. Britton 
and Atkinson (1938) often observed spontaneous variations in body 
temperature of Bradypus with no apparent reason. In the light of 
this it seems of interest that Irving et al. (1942) found that the 
resting metabolism could be readily depressed, particularly in 
relation to disturbances of the breathing pattern. The two-toed 
sloth (Gholoepus hoffman ni) is more active than Bradypus and 



shows also a higher average body temperature. At air temperatures 
of 28° C to 23 C, Bradypus showed an average of 33 C, whereas 
Choloepus showed 34.4 C . The difference in intramuscular tempera- 
tures was even greater with values of 34.6 C for Choloepus and 
32.4° C for Bradypus. Bradypus occurs only in the lower neo-tropi- 
cal altitudes and is particularly prevalent in regions with small 
fluctuations in air temperature and dense vegetation providing ample 
shade. Choloepus has a similar habitat but is also able to withstand 
colder "areas with occasional freezing in altitudes up to 7,000 or 
8,000 feet (Britton, 1941). Choloepus has not yet been studied in 
regard to temperature regulationinthesecolder areas; this project, 
however, seems to promise a great deal. The body temperatures 
of Bradypus drop precipitously when exposed to 10 G air tempera- 
ture and reached 20° C after about 5 hours. Below this temperature, 
a lethargic condition seemed to ensue. Gold was apparently a strong 
stimulus to muscular activity. Marked hypertonus was noticeable, 
but no shivering was visible at any temperature. This is of particu- 
lar interest in light of the extremely low mass of skeletal muscle 
in the sloth. Britton and Atkinson (1938) report that the skeletal 
muscle mass in Bradypus is only 25% of the body weight. The cor- 
responding figure for the higher mammals ranges between 45% and 
55%. Upon exposure to sun, the rectal temperature rose 2 C to 
4 C within 30 minutes, and the animals struggled vigorously to 
get free. Getting freedom, they promptly sought shade under the 
nearest tree. Irving et al. (1942), in their interesting study on res- 
piration in the sloth, mentioned an oxygen consumption about half 
of what is generally found in higher mammals , with values approach- 
ing what Martin found for the monotremes and some marsupials. 
Comparing the two sloths, thethree-toed is far inferior in its ability 
to maintain a fairly uniform level of body temperature and is help- 
lessly unable to venture into an environment outside the tropical 
stability. The unsurpassed slowness is interesting in the light of 
the small muscle mass. The low resting metabolism is shared with 
the other members of the Xenarthra and the monotremes and many 
of the marsupials. We have throughout this discussion seen that 
very early mammals became able to increase their metabolism by 
shivering, thus compensating for an increased heat loss. This me- 
chanism appeared before the ability to regulate heat loss intrinsi- 
cally. In the sloth things are seemingly different. The animal almost 



entirely lacks the potential to shiver. The three-toed sloth seems 
in his unique laziness to depend upon his exceptionally high insula- 
tion — nearly as great as that of many arctic animals. The result 
is an extremely labile body temperature which at low air tempera- 
tures shows greater fluctuations than encountered even in the mono- 
tremes. The three-toed sloth seems to be the least fitted of all mam- 
mals to withstand decreased air temperatures, a fact which I think 
illustrates the importance of metabolic compensation for mainten- 
ance of thermal balance. The other members of the Xenarthra show 
a far more advanced temperature regulation. The armadillos are 
very versatile animals with large distribution areas, as explicit in 
their temperature- regulating capacity. More or less casual observa- 
tions of body temperature of armadillos were made and reported 
by de Almeida and de A.Fialho (1924). Eisentraut (1932) recorded 
an average body temperature in Tolypeutes conurus of 32 C in a 
tropical habitat. Wislocki (1933), studying the nine-banded armadillo 
in Panama, reported 34.5 C as an average body temperature at air 
temperatures around 25 G.I have recently done a study on tem- 
perature regulation in this same species and will submit some of 
these data in a little more detail. 

The material studied came from Texas and not from tropical 
Central America where most of the earlier observations on this 
species have been made. I mentioned earlier that the dispersion 
of animal life north and south from the tropics is a factor presum- 
ably of great importance in the evolvement of more refined homeo- 
thermic adjustments. It struck me that the nine-banded armadillo 
is one of very few animals that incur time has taken steps to leave 
the tropical stability. The distribution area for the nine-banded 
armadillo is today very large. It ranges south into Northern 
Argentina, spreads over all of the countries east of the Andes, 
reaches the Pacific Coast in Ecuador, and extends throughout 
Central America and most of Mexico. About a hundred years ago, 
the animal crossed the border into the United States and is now well 
established in most of Texas, the southern part of Arkansas and 
Oklahoma, most of Louisiana, and southwestern Mississippi. Fur- 
thermore, it is established in Alabama and Florida. In Florida, 
the armadillo is reported to have increased its range by about 50% 
from 1954 to 1958. There are also persistent reports that there are 



armadillos in Kansas, spreading rapidly northward. At present the 
animal is still spreading north and east in the United States, Accord- 
ing to Talmage and Buchanan (1954), the migration of the armadillo 
is one of the most amazing in the animal kingdom, comparable 
almost to the lemming migrations. From the distribution area it is 
apparent that the armadillo faces very diversified ecological situa- 
tions and that its success must at least partly be dependent upon 
an extraordinary ecological potential, while the other members of 
the order, the sloths and an teaters, are still confined to their tropi- 
cal habitat. The entire range of the armadillo is characterized by 
having neither extreme cold nor extended periods of cold weather. 
The northern parts of the area, however, occasionally show quite 
low temperatures for short periods and have atypical seasonal and 
diurnal periodicity. This unique dispersion rate, bringing the animal 
out of the tropical stability and into the periodicity of seasons and 
larger diurnal variations, plus its phylogenetic position, suggested 
to me that some valuable information in regard to the phylogenetic 
development of homeothermy could be expected. 

The diurnal cycle of deep rectal temperature under controlled 
conditions is presented in Figure 9 . At a constant air temperature 
of 25 C, the diurnal cycle ranges between 34.0 C and 36.4 C. 
The animals were free to build nests from dry hay and were con- 
fined in a room of considerable size allowing for exercise. The 
diurnal cycle is obviously related to their nocturnal activity pattern. 
Following forced exercise, the deep body temperature may increase 
to 37 G or 38 G. At 30 G ambient temperature the animals were 
usually very sedate, no signs of discomfort, and rectal and skin 
temperatures were nearly constant, the difference between the two 
being surprisingly small. A rectal temperature of 34.5 C could 
correspond to a skin temperature on the soft belly as high as 34.2 C 
and on the armor 33.9 G. Under these conditions the gradient is 
less than 1 G between the core and the shell. The oxygen uptake 
showed only small variations and ranged from 200 cc to 275 cc of 
oxygen per kilo animal per hour. Figures 10 and 11 demonstrate 
these points; a slight lowering of the air temperature brought about 
a dramatic response. The most conspicuous feature was a sudden 

increase in rectal temperature. Thus, if the air temperature was 

o o 

decreased from 30 G to 25 C, the deep rectal temperature might 





oc9 o o 

O (D 
O O O O 

g oo o 

Oo o 


o o 

) 08 °( 

o o c 

^ 00° 

Figure 9. Diurnal cycle of deep rectal temperature in the armadillo. (Johansen, 1961). 



.2 ^ 









bO Ji 








































0; IOOOh 

§ 800- 


o„ o 


^ 600 











Figure 10. Oxygen consumption and deep rectal and skin temperature at various 
ambient temperature levels in the armadillo. (Johansen, 1961). 




(hard skin) 


Belly side 
(soft skin) 

Figure 11. Schematic representation of temperature gradients between core and 
dorsal and ventral sides of armadillos at different air temperatures. (Johansen, 



increase from 34.3 G to 35.4 C in less than 30 minutes. This 
increased core temperature seemed to result initially from a 
decrease in heat loss brought about by a vasoconstriction in the 
body surface. Then 15 to 20 minutes after the onset of vasoconstric- 
tion the oxygen uptake increased. The increase in oxygen consump- 
tion was correlated with the start of shivering, which apparently 
can be evoked in armadillos by the slightest stimulus. In fact, 
shivering and an attendant increase in body temperature were 
occasionally observed when the room temperature was as high as 
30 G. When exposed to cold, the animal immediately arose and 
tucked his head under his belly. His posture was ball-like and only 
a very small portion of the soft skin on the ventral side was directly 
exposed to the cold air (Fig. 9B). The increase in body temperature 
upon a decrease in air temperature did not occur only with the first 
cold stimulus. On the contrary, each time the temperature decreased 
the animal's body temperature rose immediately (Figs. 10 and 11). 
Even when the air temperature was decreased from G to -10 C, 
and the animal already had a very high heat production, the body 
temperature rose. On one occasion when the change was from C 
to -6 C, the rectal temperature increased from 35.7 G to 36.1 G 
in less than 10 minutes. During the stepwise decrease of ambient 
temperature from 30 G to -10 G in a period of 5 hours, the total 
increase in rectal temperature was 3.5 G (Fig. 10) and the body 
temperature on the armor decreased about 10 G to 23 G. These 
temperature changes occurred in a step- like pattern following the 
changes in air temperature. Upon a sudden decrease in air tempera- 
ture from 30 G to G, the rectal temperature rose 2 G in 40 
minutes. But this increasewas not as great as was the total response 
to step- reductions from 30° G to 20 C to 10 G to G. This very 
conspicuous increase in rectal temperature upon cold exposure has 
to my knowledge not been reported before. It may have escaped 
notice because rectal temperatures were not continuously recorded. 
Bartholomew, in his paper on Setonix (1956) makes the statement 
that at -10 G, the animal kept an elevated body temperature for 
quite some time during the exposure period. Meanwhile, he did not 
persuade this point further in his discussion. In my interpretation, 
this fact, so very explicitly demonstrated in the armadillo, shows 
that the effector mechanisms that modify heat loss, like vasocon- 
striction, etc., operate very promptly but entirely out of pace with 




^ 35- 



^ 30- 


^ 25- 

^ - 30° 
^ 0° 



4 5 


Figure 12. Continuous records of deep rectal temperatures of the armadillo 
at various air- temperature levels. (Johansen, 1961). 

the governing thermostatic control which fails in the armadillo 
with several degrees. You may rightly ask how long this unsteady 
condition goes on. P^igure 12 demonstrates the events during long- 
time exposure to various ambient temperatures. When an experi- 
ment began at 30 C, the deep rectal temperature fluctuated 

o o 

between 33.9 C and 34.5 G, as can be seen in Figure 12. The 

animals were quiet and relaxed. Occasionally they walked around, 
but the activity was always transient and on a low level. The ani- 
mals responded to an air temperature of C with a rapid increase 
in body temperature like that recorded in the metabolic studies. 
This response was accompanied by curling into a ball as described 
earlier, and shivering began immediately. These bodily responses 
continued throughout the test. At the end the animal showed no 
signs of exhaustion. The curve for C in Figure 12 illustrates 
the typical variation in deep rectal temperature. It rose rapidly 
to 36.3 G in this animal within the first half hour. After an hour 
the temperature dropped slightly to 35.5 G and remained between 
35 C and 35.5 G for more than 6 hours. Toward the end of the 



exposure, the temperature dropped slightly; the last temperature 
recorded in this animal was 34 C. In similar experiments at 
-10 C, the initial increase in rectal temperature was even more 
rapid and pronounced (Fig. 12). After slightly more than an hour, 
the rectal temperature had declined to the level recorded at the 
beginning of the experiment. For the next 2 1/2 hours the tempera- 
ture fluctuated around 35 G. Then suddenly the animal began to 

run about the room scratching at the walls, the rectal temperature 

started to drop, and within five hours it had fallen to 25.5 C. At 

this low body temperature, the animal showed no signs of severe 
fatigue and was amazingly coordinated. Its movements were harmon- 
ized and it eagerly drank some milk which it was offered. The highly 
organized behavior at such a low body temperature seems to be 
unique among non- hibernating mammals and is probably of great 
functional value to the armadillo. Although the pattern of -10 C 
shown in Figure 12 is typical, the length of time the body tempera- 
ture was maintained near 35 C varied from animal to animal, being 
less than 2 l/2 hours in some and as long as 5 l/2 hours in others. 
The period of relatively constant temperature terminated when the 
animal abandoned the protective posture, which no doubt provided 
good insulation. It should be emphasized that this kind of exposure 
does not at all give a true picture of the armadillo's tolerance to 
severe cold in natural environments. It will be pointed out later that 
the armadillo is very much concerned with and dependent upon nest 
building and social habits for survival at low ambient temperatures. 

I should like to spend a few minutes commenting upon the intact- 
ness of coordinated behavior at drastically low body temperatures. 
Other workers have mentioned the fact that normal behavior per- 
sists at very low body temperatures in many primitive animals. 
This fact has, however, never been appropriately evaluated. It 
strikes me since it is a common characteristic among most of the 
lower mammals studied in this respect. Comparing this with the 
lethargic condition and the concurrent impairment of coordination 
that appear in higher mammals, for example carnivores and 

ungulates at temperatures which are for them only slightly hypo- 

o o 

thermic (30 G to 33 G), it seems of the highest significance. I 

think the persistence of coordinated behavior at low body tempera- 
tures in primitive forms has great survival value and greatly 



increases their independence from the environment. I think, fur- 
thermore, that this is an important point in the phylogeny of homeo- 
thermy and more than anything that it should be subjected to further 

My data on the tolerance to cold in a long-time experiment 
differs markedly from apparent results of earlier investigations 
on armadillos. The conspicuous increase in rectal temperatures 
has not been reported previously and it may have escaped notice 
because the rectal temperatures during cold exposure were not 
continuously recorded. Wislocki (1933) observed the following rectal 
temperature patterns in armadillos transferred from 28 C to 
C air temperature: at time, 35 C; at 1-hour exposure, 34 C; 
at 2- hour exposure, 31.5 C; at 3- hour exposure, 30 C. Recently, 
Enger (1957) reported that the armadillo, opposum, and three- 
toed sloth are poor thermal regulators and lose body heat to a 
considerable extent during a cold stress, that is, 4 to 6 hours at 
10° C air temperature. These results differ basically from those 
of the present study. It would be highly desirable to determine 
whether this difference stems from the fact that Wislocki 's and 
Enger's measurements were made on animals from a strictly 
tropical habitat, Barro Colorado, in Panama, whereas the subjects 
of this study came from Texas. This would be a most notable 
demonstration of adaptation to cold. 

The increase in oxygen consumption upon lowering of the air 
temperature for a representative experiment is shown in Figure 
10, and for all experiments in Figure 13. There was more than 
a five-fold increase in oxygen uptake when the air temperature 
was reduced from 30 C to -10 C. The relatively great spread 
in the data at the lower temperatures is probably related to the 
differences in the rectal temperatures of different animals. No 
attempt was made to test responses to temperatures below -10 C 
because some animals try to escape exposure to this temperature. 
These activities of course involved the abandonment of the protec- 
tive posture and large losses of heat occurred. Ordinarily the 
animals showed an amazing ability to remain crouched in their 

ball- like posture for hours when the temperature was as low as 





Figure 13. Oxygen consumption versus ambient temperature for the armadillo. 
(Johansen, 1961). 

In the light of the observations on the metabolic response to 
cold, I find it proper to discuss briefly the concept we all have been 
repeatedly informed about, namely that the effector mechanisms 
affecting heat dissipation and conser-vation, what we call physical 
temperature regulation, come first into play and reach their effect- 
ive limits before metabolic compensation sets in. In other words, 
at the point of the critical temperature, the ability for physical 
regulation is exhausted. This fact has been repeatedly stated from 
studies on temperate and arctic mammals and has received value 
as a concept in our understanding of temperature regulation. I 
can see no reason why this strict sequence has to have general 
value. Martin's data on the platypus and the marsupials indicate 
a gradual, simultaneous time of action for both processes. My 
own data on the armadillo likewise support the idea that maximal 
insulation, including vasomotor adjustment, does not necessarily 
need to reach the end point before metabolic compensation sets 
in. The sequential arrangement has, of course, an obvious biologi- 
cal rationale, by saving fuel. However, in line with my earlier 



reasoning I feel that such a strict sequential play-off of the com- 
pensatory mechanisms will require a rather precise central ner- 
vous thermostatic control which in many forms may not have 
reached the perfection needed. 

For the metabolic rate of the armadillo, my studies showed a 
low resting value of about 250 cc/(kilo x hr) at an air tempera- 
ture of 30 C. These measurements are in accord with those 
reported by Scholander et al. (1943). They found values averag- 
ing 180 cc/(kilo xhr) and varying between 150 and 280. These values 
are roughly half the size of those reported by de Almeida and 
de A. Fialho (1924). However, Scholander etal. felt that the experi- 
mental approach may have been a factor in the recording of the 
high values by these authors. The resting metabolism of the arma- 
dillo, then, is obviously less than that common in mammals of the 
same size. It is slightly more than half that simultaneously measured 
in rabbits of comparable size. The measured metabolic rate in 
rabbits corresponded with values given by Benedict in 1938. The 
resting metabolism of the armadillo, although somewhat higher, 
approaches that found in sloths by Irving etal. (1942). Martin (1903) 
reported an air temperature of 30 C coinciding with the lowest 
metabolic value for the monotremes, Echidna and Omithorhynchus , 
and for several marsupials. Throughout the whole temperature range 
studied, Martin found the metabolism to be lower in monotremes and 
marsupials than in higher mammals like cats and rats. The ques- 
tion arises: Is this low level of resting metabolism related to the 
low resting body temperatures found in these species ? The results 
of my study on the armadillo suggest that this may be partly true, 
but that differences in the thermostatic mechanisms are also impor- 
tant factors in the dissimilarity between the armadillo and more 
advanced homeotherms. For the same reason critical temperatures 
and critical gradients are more or less meaningless when applied 
to these species. As mentioned, the armadillo will occasionally 
start shivering and increase its body temperature at an air tempera- 
ture of 30 G. The word "critical," as in critical temperature and 
critical gradient, is obviously not pertinent to this situation. Martin 
argued similarly that the low metabolism in monotremes and mar- 
supials can result from the following factors: a greatly diminished 
heat loss, a lower body temperature level, and failure to maintain a 
constant body temperature. 



The increase in heat production when air temperature declines 
has several noteworthy features. The steepness of the slope of 
metabolism versus air temperature is fairly great and of about 
the same magnitude as for the two-toed sloth. The fact that the 
naked armadillo and the hairy sloth with about the same resting 
metabolism showed similar regressions in this respect supports 
the idea that the armadillo has a potent vasoconstrictor ability. 
The fact that the armadillo showed a step- wise increase in insula- 
tion may, however, somewhat invalidate comparison of it to the 
sloth. The peak metabolic values in the armadillo vary from five 
to six times larger than the resting values. Such high metabolic 
rates were recorded for as long as 6 hours, which was the longest 
time oxygen consumption was measured at one fixed low tempera- 
ture. According to Scholander et al. (1950) the maximum increase 
in heat production is time-dependent and seldom more than four 
times the resting value. They state that this relation is valid for 
long-time experiments, but they do not define "long time." In the 
armadillo the increased heat production results from a progressive 
augumentation of shivering. The shivering pattern is closely related 
to the crouched posture and the armadillo rarely utilizes moving 
about as a means for increasing heat production in the cold. 
Occasionally when they abandoned the immobile posture, the ani- 
mals experienced great heat losses. To demonstrate more expli- 
citly the difference in metabolism and response of body temperature 
to cold exposure, I did some experiments with rabbits. In Figure 
14, we can see that deep rectal temperatures and oxygen consump- 
tion were followed at various ambient temperature levels ranging 
o o o CO 

from +30 Gto-6 G.AtSO C air temperature, the rectal tempera- 
tures of the armadillos are more than 4 C lower than for the rab- 
bits. Upon a gradual step-wise decrease in air temperature, the 
armadillos show an increasing body temperature, accompanied by 
a steep increase in oxygen consumption, whereas the rabbits' con- 
dition is unchanged. The procedure lasted 4 to 5 hours. These studies 
demonstrate a conspicuous over-compensation to the cold stress 
in the armadillo. Inother words, the body temperature is drastically 
raised at the expense of an increased metabolism. The fact that this 
over-compensation takes place so rapidly and to such a large extent 




a. 800 

600 H 


o o 


C 34- 



O O 



5 10 15 20 



Figure 14. Comparison between oxygen consumption and deep rectal tempera- 
tures for armadillos smd rabbits at different ambient temperature levels. Open 
circles indicate rabbit data; crosses and regression line indicate the results 
from the armadillo experiments. (Johansen, 1961). 



demonstrated the presence of well-developed mechanisms for 
compensatory heat production. However, this heat production and 
conservation are not governed by the same thermostatic arrange- 
ments that are present in rabbits. 

Grant me also a few words about insulation in the naked arma- 
dillo. The very large temperature gradients in the extremities and 
snout shortly after the beginningof cold exposure are likely of great 
significance (Figs. 15, 16, and 17). It is interesting to compare this 
observation with those of Scholander and Krog on the sloths from 
1957. These investigators suggest that the vascular bundles rete 
mirabile in the limbs of the sloths strongly facilitate the conserva- 
tion of central body heat at the expense of a profound cooling of the 
limbs. The principle involved is thought to be a counter-current 
heat exchange in the vascular bundles which provide a greatly 
enlarged contact surface between the counter-streaming arteries 
and veins (Scholander and Schevill, 1955; Scholander, 1958). In 
the sloth, subcutaneous temperature gradients in the legs were as 

Figure 15. Cutaneous temperatures at various sites on the armadillo's body 
at air temperatures of 25 C - 30 C (upper numbers) and 5° C (lower numbers). 
(Johansen, 1961). 



Figure 16. Cutaneous temperatures at various sites on the armadillo's body at 
air temperatures of 25 C - 30 C (upper numbers) and 5 C (lower numbers). 
(Johansen, 1961). 








»^ rv) ro OJ OJ 

00 r>o o> O -f^ 

__. I . \ ■ I I i_ 

Figure 17. Intramuscular an(i subcutaneous temperature gradients along the 
hind extremity of an armadillo. Open circles and squares show subcutaneous 
gradients; filled circles and squares, intramuscular gradients. Circles are record- 
ings taken at air temperature of 25 C; squares at C (Johansen, 1961). 



o o 

great as 0.2 G per cm at ambient temperatures of 25 C to 

27 C. Such gradients are 10 times steeper than those in the 
human arm under similar conditions. The same vascular struc- 
tures are present iji the limbs of other living Xenarthra, the 
anteaters and the armadillos. The present study, showing very 
large temperature gradients at low air temperatures, was not 
designed to verify the counter- current hypothesis and can give 
no conclusive evidence in this respect. 

We note from the important works of Irving and Scholander 
and associates that the insulative value of animals generally 
increases as we proceed north and south to the arctic and ant- 
arctic regions. The almost unbelievable insulation attained by 
some of the larger arctic mammals, like the husky, the wolf, 

and the fox, makes possible a maintained resting metabolism down 

o o 

to 60 G to 70 G below zero. This fact, I think, poses an inter- 
esting question: How do these animals get rid of the excess heat 
produced during the extensive exercise they necessarily have to 
practice? You may think that panting is enough to keep the body 
temperature at so-called normal levels of 37 G to 38 G. I have 
done this winter some measurements on the exercising husky 
sled dogs. The results show that shortly after the start of exer- 
cise, both intramuscular and deep rectal temperatures reach 

o o 

levels of about 41 G to 42 G in an air temperature down to 

-50 G. These surprisingly high temperatures did not in any way 
impair the performance of the dogs, which could keep on working 
for 6 to 8 hours at the same speed, allowing only brief periods 
of rest. Studies done on dogs from warmer temperate regions 
indicate an impairment of function and heat collapse at lower body 
temperatures, down around 41 G. Although few data are availa- 
ble so far, it seems probable that the unbelievably great insula- 
tion of some of the arctic mammals has resulted in an adaptive 
tolerance to an elevated body temperature. At least the husky 
provides information in this direction. My studies on the heavily 
insulated muskrat indicate an entirely different solution to the 
problem. Time seems, however, to prevent us from going into 



Let us next look at the physiological responses to high air tem- 
peratures in the armadillo. Figure 18 shows the changes in rectal 
and skin temperatures and oxygen consumption when the ambient 
temperature was increased above 30 G. The relative humidity was 
kept below 30%. As the air temperature was increased in steps to 
42 C over 4 to 5 hours, the rectal temperature rose to 40 C. If 
the air temperature was increased from 30 Gto 42 C in one step, 
the rectal temperature reached 40 C in less than 3 hours. From 

the behavior of the animals, 40 C seems near the upper lethal 

limit for the rectal temperature, although 41.5 C was repeatedly 

tolerated for periods of less than 1 hour. The skin temperature on 
the back armor rose markedly in a step- wise fashion following the 
changes in room temperature. The skin remained completely dry, 
however, and no active sweat glands were detected with either of 
the two methods used. The circulation to the skin was greatly aug- 
mented; even the dorsal armor blushed pinkish red. The ears were 
markedly vasodilated, constantly vibrating. At high air temperature 
panting is an important avenue of heat loss for the armadillo. The 
respiratory rate rose from 30 to 40 breaths a minute to almost 200 
a minute. The g-reatest increase in rate seemed to occur when the 


rectal temperature was between 37 C and 28 C (Fig. 19). During 
panting the nostrils were red and vibrating intensely. There seemed 
also to be a rather profuse salivation from the buccal mucosa. How- 
ever, no licking ever occurred. The oxygen consumption rose from 
about 240 cc/(kilo x hr) to 400 cc/(kilo x hr). Part of this increase 
was related to the muscular activity of hyperventilation, part to 
the elevated rectal temperature. The significance of panting in the 
armadillo's response to heat was demonstrated by anesthetizing 
an animal when its body temperature was about 38 C and the air 
temperature was 40 G (Fig. 19). As soon as the respiratory rate 
was depressed to normal or subnormal levels, the rectal tempera- 
ture rose sharply. During exposure to the hot environment, the 
animals usually turned on their sides immediately, stretching their 
front legs forward and their hind legs backward so that their ven- 
tral surface was maximally exposed. When the rectal temperature 
was 40 G, the animals were obviously uncomfortable and some 
attempted to escape. 



' 400 ^ 

O O O O O o 




2H 3H 

Figure 18. Oxygen consumption and deep rectal and skin temperatures at 
various ambient-temperature levels in the armadillo. (Johansen, 1961). 





01 f^ 


5? " 

3 ^ 






Next, I have a few closing words about the role of behavior. 
The almost unbelievable achievement in the temperature regula- 
tion of reptiles, reached solely by behavioral means, ought to be 
a strong reminder to all of us that we can also reasonably expect 
behavior to play a crucial role in temperature regulation in mam- 
mals. This I am sure we all agree upon, but in our laboratory 
experiments we necessarily have to dispense with most of the sub- 
jects' opportunities for behavioral regulation. Our knowledge 
regarding the role of behavior in the physiological phylogeny of 
temperature regulation is therefore very limited. In the monotremes 
it is of paramount importance. In the primitive Echidna behavior 
seems to be the only way heat dissipation can be effected and is 
thus of vital consequence, particularly on the hot side. To evaluate 
the importance of behavior along the phylum is virtually impos- 
sible, and I will make no attempt to do so. With the increasing 
development of the cerebral capacity, behavior may reach ultimate 
sophistication in man with his air-conditioned houses. Physiolo- 
gists and anthropologists have thus taught us that man's invasion 
of the climatic extremes, such as the Eskimos, Lapps, and others, 
is almost entirely achieved by behavior, with rather subtle changes 
in physiological adaptation. Since we are discussing evolution, it 
may be of interest thatbehavior may also show regression as a fac- 
tor in human temperature regulation. I am referring to man's or 
more correctly, woman's vanity, explicit in sheer nylon stockings 
in 40 G below weather, readily observable in the streets of 
Fairbanks every winter. 

An important part of the armadillo's temperature regulating 
ability is represented by its behavioral pattern. Thus, there is no 
doubt that the ball- like posture is an extremely important means 
of increasing the insulation. Measurements made by Buttner (1938) 
demonstrate a reduction of 50% in the surface area when a man 
curls into ball- like posture. The importance of posture was shown 
clearly for the armadillo as well by the skin temperature measure- 
ments and the experiments with the heat- flow discs. The building 
of nests is also a most important factor in survival during exposure. 
Without overstating it, nest- building seems to be a highly developed 
social habit which is of the greatest functional significance in the 
survival and expansion of the species. It is of particular interest in 



this connection, as reported by Scho lander et al. (1950) that some 
of the smaller arctic mammals with furry insulation like that of 
tropical mammals are dependent upon burrows and nest-building 
for their survival. Notably, measurements made by Scho lander et al. 
showed that the insulative value of a lemming nest is roughly 1.5 
times that of the lemming fur. If the nest is covered with snow, its 
insulative value would presumably be even greater. In considering 
the evolution of homeothermism, then, one must include behavioral 
patterns as essential and indispensable parts of the whole system. 

The rather loosely connected information I have given you may 
qualify for a tentative outline of one probable way the evolutionary 
sequence of homeothermy has taken place. Figure 20 is an attempt 
to put the factors together, but only in a qualitative way, since 
measurements are lacking in most parts. The first successful 
efforts to maintain a fairly uniform level of body temperature began 
on the psycho-physiologicalorbehavioral level. Such an achievement 
would necessitate a well-developed sensory system for temperature, 
as well as a nervous coordination of the effector mechanism, 



^^ ^yC^^x^^^y^y^-^-^^y'^y^y^^^^-'^y^^ 


fgm/x stnsotion and CMS inttgrvtiofi of responses to the environment 

CNS th«rmostotlc control of chemieol and physical tamp regulation 

Figure 20. Simplified schematical drawing of a possible route for the evolu- 
tionary sequence of homeothermy in mammals. (Johansen, 1961). 



like locomotion, etc. The first directly physiological factor brought 
into the picture of homeothermy seems to have been a variation in 
metabolic heat production, thus for the first time releasing animal 
life somewhat from the environment. As discussed above such regu- 
lation has serious limitations, and collapse occurs at severely high 
air temperatures and outside the tropical stability. The next step 
toward advanced homeothermy seems to have been the appearance 
and development of the regulation of physical he at exchange. We can 
trace a gradual improvement of such function along the phylum but 
also closely correlated to the thermal stress imposed by the envi- 
ronment. We know from the important works of Irving and Scho lander 
and associates that under the extreme conditions confronting the 
arctic mammals, homeothermy is first of all accomplished and 
maintained by adjustment of the shell-core temperature gradient— 
or in other words, by adjustments of the insulation. The efficient 
regulation of temperature under changing conditions in the environ- 
ment must ultimately be entirely dependent upon an integrated con- 
trol of heat loss and heat exchange by thermostatic arrangements. 
These thermostatic arrangements have reasonably developed grad- 
ually becoming increasingly complex and accurate. The data pro- 
vided by Bartholomew on Setonix , as well as my own data on the 
armadillo, I think, provides an example of animals whose tempera- 
ture regulating ability has reached a stage where the degree of 
thermostatic control is a factor limiting the efficiency with which 
the animals maintain homeothermy. Efficient central nervous 
thermostatic control seems thus to have been the last factor devel- 
oped to perfection in homeothermy. 

If you will grant me another minute, I will admit that there is 
an obvious trend in the lower forms of mammals for body tempera- 
ture to be characterized by a large range and a lower set average. 
In all treatments on the subject, however, this fact is regarded as 
typifying primitive forms and thus the evolution of homeothermy. 
It has been stated that the least variable factor in the whole pic- 
ture of homeothermy is the body temperature. From my own data, 
as well as other information accumulated recently, I am strongly 
opposed to this view. In my mind it is entirely conceivable that a 



large activity range of body temperature may express a specializa- 
tion rather than a primitive condition. A truly fixed body tempera- 
ture, fluctuating within very narrow limits, would for many species 
be highly uneconomical, or not at all obtainable in the special envi- 
ronments they face. This viewpoint gets support from my work on 
the birch mouse, a rodent which shows diurnal fluctuations in body 
temperature up to 20 C in the summer time (Johansen and Krog, 
1959). The large activity range of the husky gives additional support. 
The work done on the camel by Schmidt- Nielsen et al. (1957) and 
on the rhinoceros by Albrook et al. (1958), and practically all the 
smaller animals subjected to thoroughstudy recently show the same 
thing (Morrison and Ryser, 1959). Let us not forget Joseph 
Barcroft's words (1934) that every adaptation is an integration. Let 
us remember there is more to it than just keeping a constant core 
temperature. The effector systems involved in temperature regula- 
tion ^have other tasks to perform, which is so strikingly apparent 
from Schmidt- Nielsen's study on the came LI am confident that when 
our knowledge of body temperature ranges during activity and other 
bodily performances, as well as sleep, is extended, this will dis- 
close a larger range of body temperatures than we are familiar with. 
Let us not a priori let such a large temperature range be classified 
as a primitive sign. Moreover, I feel from my work on the arma- 
dillo that the specializations 1 have mentioned above can also develop 
in forms of lower phylogenetic ranking and thus complicate, and 
maybe somewhat invalidate, the body temperature range as a clear 
measure of the phylogenetic standing. I would like to submit this idea 
as a challenge to present concepts of temperature regulation. 

Ed. note: The discussion of Dr. Johansen's paper was postponed until the 
following session due to lack of time. It is incorporated with the discussion of 
Dr. Irving's paper. 



1. Allbrook, D. B., A. M. Harthoom, C.P. Luck and P. G. Wright. 
1958. Temperature regulation in the white rhinoceros. J. 
Physiol. 143:51. 

2 Almeida, O. de, and B. de A. Fialho. 1924. Temperature et 
metabolisme du tatou. Compt. rend. Soc. biol. 90:734. 

3. Barcroft, J. 1934. Features in the Architecture of Physiological 

Function. Cambridge Univ. Press. 368 pp. 

4. Bartholomew, G. A. 1956. Temperature regulation in the macro- 

pod marsupial, Setonix brachyrus . Physiol. Zool. 29:26-40. 

5. Bartholomew, G. A. and T. J. Cade. 1957. Temperature regula- 

tion, hibernation and aestivation in the little pocket mouse, 
Perognathus longimembris . J. Mammal. 38:60-72. 

6. Benedict, F. G. 1938. Vital Energetics. Carnegie Inst, of Wash. 

PubL No. 503. 215 pp. 

7. Bogert, G. M. 1949. Thermoregulation in reptiles, a factor in 

evolution. Evolution. 3:195-211. 

8. BoUiger, A. and M. H. Hardy. 1944. The sternal integument of 

Trichosurus vulpecula . J. Proc. Roy. Soc. N. S. W. 78:122- 

9. Britton, S. W. 1941. Form and function in the sloth. Quart. Rev. 

Biol. 16:13-34 and 190-207. 

10. Britton, S. W. and W. E. Atkinson. 1938. Poikilothermism in 

the sloth. J. Mammal. 19:94-99. 

11. Burrell, H. 1927. The Platypus. Sydney, Angus and Robertson, 

Ltd. 277 pp. 



12. Buttner, K. 1938. Physikalische Bioklimatologi. Leipzig. 

13. Cowles, R. B. and C.M.Bogert. 1944. A preliminary study of the 

thermal requirements of desert reptiles. Bull. Am. Mus. Nat. 
Hist. 83:265-296. 

14. Darlington, P. J. 1948. Geographical distribution of cold blooded 

vertebrates. Quart. Rev. Biol. 23:105-23. 

15. Eisentraut, M. 1932-33. Biologische Studien im Bolivianischen 

Chaco. IV. Die Warme Regulation beim Kugelgurteltier 
(Tolypeutes conurus J. S. Geoff.). Zeitschr. Vergleich. 
Physiol. 18:174-85. 

16. Eisentraut, M. 1955. Apropos de la Temperature de quelques 

mammiferes de type primitif. Mammalia. 19:437-443. 

17. Eisentraut, M. 1956a. Der Winterschlaf mitseinenokologischen 

und physiologischen Begleiterscheinungen. Gustav Fischer 
Jena. 160 pp. 

18. Eisentraut, M. 1956b. Temperaturschwankungen bei niederen 

Saugetieren. Zeitschrift fur Saugetierkunde. 21:49-52. 

19. Enders, R. K. and D. E.Davis. 1936. Body temperature of some 

Central American mammals. J. Mammal. 17:165. 

20. Enger, P. S. 1957. Heat regulation and metabolism in some 

tropical mammals and birds. Acta Physiol. Scandinavia. 

21. Herter, K. 1934. Korpertemperatur und aktivitet beim Igel. 

Zeitschr. f. vergl. physiol. 30:511-544. 

22. Hesse, R. 1937. Ecological Animal Geography. John Wiley and 

Sons. 597 pp. 

23. Higginbotham, A. C. and W. E. Koon. 1955. Temperature regula- 

tion in the Virginia opposum. Am. J. Physiol. 181:69-71. 



24. Irving, L., P. F. Scholander, and S. W. Grinell. 1942. Experi- 
mental studies of the respiration of sloths. J. Cell, and 
Gomp. Physiol. 20:189-210. 

25. Irving, L. and J. Krog. 1955. Temperature of skin in the Arctic 

as a regulator of heat. J. Appl. Physiol. 7:355-66. 

26. Johansen, K. and J. Krog. 19 59. Diurnal body temperature 

variations and hibernation in the birch mouse Si cista 
betulina . Am. J. Physiol. 196:1200. 

27. Johansen, K. 1961. Temperature regulation in the ninebanded 

armadillo (Dasypus novemcinctus mexicanus ) . Physiol. Zool. 

28. Kredel, F. E. 1928. Notes on the temperature of the sloth. J. 

Mammal. 1:48-51. 

29. Krieg, M. 1952. Der Kaltestarre Kolibri. Naturw. Rdsch. 5, 

S. 25-26. 

30. Lendenfeki, R. von. 1886. Zur Brutpflege von Echidna. Zooli- 

gischer Anzeiger 9: Jahrgang. 9 pp. 

31. Miklouho-Maclay, N. de. 1883. Temperature of the body of 

Echidna hystrix. Proc. Linn. Soc. New South Wales. 8:425. 

32. Martin,G. J. 1903. Thermal adjustment and respiratory exchange 

in monotremes and marsupials. A study in the development 
of homeothermism. Phil. Trans. Roy. Soc. London B. 

33. Martin, G. J. 1930. Thermal adjustment of man and animals to 

external conditions. Lancet. 2:561-567, 617-620, 673-678. 

34. Morrison, P. R. 1945. Acquired homeothermism in the preg- 

nant sloth. J. Mammal. 27:272-75. 



35. Morrison, P. R. 1956. Temperature regulation in three Central 

American mammals. J. Cell, and Comp. Physiol. 27:125-137. 

36. Morrison, P. R. 1957. Observations on body temperature in a 

hedgehog. J. Mammal. 38:254-255. 

37. Morrison, P. R. 1959. Body temperature in some Australian 

mammals. I. Chiroptera. Biol. Bull. 116:484-497. 

38. Morrison, P. R. and F. A. Ryser. 1959. Body temperature in 

the white- footed mouse, Peromyscus leucopus noveboracen - 
sis . Physiol. Zool. 32:90-103. 

39. Morrison, P. R., F. A. Ryser and A. R. Dawe. 1959. Studies on 

the physiology of the masked shrew Sorex ciner eus . Physiol. 
Zool. 32:256-271. 

40. Rand, A. L. 1925. On the habits of some Madagascar mammals. 

J. Mammal. 16:89-104. 

41. Robinson, Kathleen W. 1954. Heat tolerances of Australian 

monotremes and marsupials. Austr. J. Biol. Sci. 7:348-360. 

42. Robinson, K. and P. R. Morrison. 1957. The reaction to hot 

atmospheres of various species of Australian marsupial and 
placental animals. J. Cell, and Comp. Physiol. 49:455-478. 

43. Schmidt- Nielsen, K., B. Schmidt-Nielsen, S. A. Jamum, and 

T. R. Houpt. 1957. Body temperature of the camel and its 
relation to water economy. Am. J. Physiol. 188:103-12. 

44. Scholander, P. F., L. Irving and S. W. Grinell. 1943. Respira- 

tion in the armadillo with possible implications as to its bur- 
rowing. J. Cell, and Comp. Physiol. 21:53-63. 

45. Scholander, P. F., R. J. Hock, V. Walters, F. Johnson and L. 

Irving. 1950. Heat regulation in some arctic and tropical 
mammals and birds. Biol. Bull. 98-99:237-58. 



46. Scholander, P. F. and W. E. Schevill. 1955. Countercurrent 

vascular heat exchange in the fins of whales. J. Appl. Physiol. 

47. Scholander, P. F. and J. Krog. 1957. Countercurrent heat ex- 

change and vascular bundles in sloths. J. Appl. Physiol. 
10 '.40 5- 11. 

48. Scholander, P. F. 1958. Countercurrent exchange; a principle 

in biology. Hvalradets skrifter. 44;5-24. 

49. Semon, R. 1894. Notizen uberdie Korpertemperaturder nieder- 

sten Saugetiere (Monotremen) . Arch. ges. Physiol. 58:229- 

50. Sutherland, A. 1897. The temperatures of reptiles, monotremes 

and marsupials. Proc. Roy. Soc. of Victoria. N. S. 8:57-67. 

51. Talmage, R. T. and G.D.Buchanan. 1954. The armadillo ( Dasy- 

pus novemcinctus) . A review of its natural history, ecology, 
anatomy and reproductive physiology. Rice Institute Pamph- 
lets. 41 (No. 2):l-35. 

52. Wardlaw, H. S. H. 1915. The temperature of Echidn a aculeata . 

Proc. Linn. Soc. N. S. Wales. 40:231-258. 

53. Winge, H. 19 41. The Interrelationship of the Mammalian Genera. 

Vol. I. Monotremata, Marsupialia, Insectivora, Chiroptera. 
Edentata, Reitzels Forlag. Kobenhavn. 412 pp. 

54. Wislocki, G. B. 1933. Location of testes and body temperature 

in mammals. Quart. Rev. Biol. 8:385-96. 

55. Wislocki, G. B. and R. K. Enders. 1935. Body temperature of 

sloths, anteaters and armadillos. J. Mammal. 16:328-329. 

56. Wood-Jones, F. 1923. The mammals of South Australia. (Govt. 

printer Adelaide).